ArticlePDF Available

Automated visual inspection system for wood defect classification using computational intelligence techniques


Abstract and Figures

This article presents improvements in the segmentation module, feature extraction module, and the classification module of a low-cost automated visual inspection (AVI) system for wood defect classification. One of the major drawbacks in the low-cost AVI system was the erroneous segmentation of clear wood regions as defects, which then introduces confusion in the classification module. To reduce this problem, we use the fuzzy min–max neural network for image segmentation (FMMIS). The FMMIS method grows boxes from a set of seed pixels, yielding ideally the minimum bounded rectangle for each defect present in the wood board image. Additional features with texture information are considered for the feature extraction module, and multi-class support vector machines are compared with multilayer perceptron neural networks in the classification module. Results using the FMMIS, additional features, and a pairwise classification support vector machine on a 550 test wood image set containing 11 defect categories show 91% of correct classification, which is significantly better than the original 75% of the low-cost AVI system. The use of computational intelligence techniques improved significantly the overall performance of the proposed automated visual inspection system for wood boards.
Content may be subject to copyright.
International Journal of Systems Science
Vol. 40, No. 2, February 2009, 163–172
Automated visual inspection system for wood defect classification using computational
intelligence techniques
Gonzalo A. Ruz
*, Pablo A. Este
and Pablo A. Ramı
Manufacturing Engineering Centre, Cardiff University, Cardiff, UK;
Department of Electrical Engineering,
Universidad de Chile, Casilla 412-3, Santiago, Chile
(Received 2 May 2006)
This article presents improvements in the segmentation module, feature extraction module, and the classification
module of a low-cost automated visual inspection (AVI) system for wood defect classification. One of the major
drawbacks in the low-cost AVI system was the erroneous segmentation of clear wood regions as defects, which
then introduces confusion in the classification module. To reduce this problem, we use the fuzzy min–max neural
network for image segmentation (FMMIS). The FMMIS method grows boxes from a set of seed pixels, yielding
ideally the minimum bounded rectangle for each defect present in the wood board image. Additional features
with texture information are considered for the feature extraction module, and multi-class support vector
machines are compared with multilayer perceptron neural networks in the classification module. Results using
the FMMIS, additional features, and a pairwise classification support vector machine on a 550 test wood image
set containing 11 defect categories show 91% of correct classification, which is significantly better than the
original 75% of the low-cost AVI system. The use of computational intelligence techniques improved
significantly the overall performance of the proposed automated visual inspection system for wood boards.
Keywords: AVI systems; image segmentation; wood defect classification; multi-class support vector machines
1. Introduction
Automated visual inspection (AVI) systems are an
automated form of quality control normally achieved
using a camera connected to a computer. The AVI
framework includes the following stages (Pham and
Alcock 2003):
(1) Image acquisition: to obtain an image of the
object to be inspected.
(2) Image enhancement: to improve the quality
of the acquired image, which facilitates later
(3) Image segmentation: to divide the image into
areas of interest and background. The result of
this stage is called the segmented image, where
objects represent the areas of interest.
(4) Feature extraction: to calculate the values of
parameters that describe each object.
(5) Classification: to determine what is represented
by each object.
AVI systems have been used with success in several
industries, achieving considerable improvements com-
pared to human inspection. Some examples are: pulp
inspection under critical lighting (Campoy, Canaval
and Pen
˜a 2005), contaminant removal from wool
(Zhang et al. 2005), inspection of colour tablets in
pharmaceutical blisters (Derganc, Likar, Bernard,
Tomazˇ evic
ˇand Pernus
ˇ2003), misplaced component
inspection for printed circuit boards (Enke and Dagli
1997), and defect detection in textile fabrics
(Abouelela, Abbas, Eldeeb, Wahdan and Nassar
2005). More industrial applications as well as impor-
tant issues and directions for designing and developing
industrial vision systems can be found in Malamas,
Petrakis, Zervakis, Petit and Legat (2003).
In the wood industry, there are human operators
that identify and locate board areas containing defects.
Errors in determining a defect and its location, known
as operator error, increases product cost. Buehlmann
and Thomas (2002) carried out a study to measure the
human operators performance in the wood defect
detection problem. Three types of operator error were
investigated: (a) when the operator marked a defect
when there was none, (b) when the operator missed
a defect, and (c) when the operator marked a defect
inside the defective area of the strip (incorrect marking
of a defect within its boundaries). Results showed that
78.2% of all defects were marked incorrectly.
Specifically, 2% of all marks made were type (a)
errors, 43.4% were type (b) errors, and 32.8% were
type (c) errors. This high error rate translated in a high
rate of rejected parts after the processing (22.0%).
*Corresponding author. Email:
ISSN 0020–7721 print/ISSN 1464–5319 online
!2009 Taylor & Francis
DOI: 10.1080/00207720802630685
Downloaded By: [Ruz, Gonzalo A.] At: 15:49 16 February 2009
They concluded that AVI systems could solve many of
the operator errors observed in this study. Due to the
high rate of human errors, such systems do not need
to be flawless. If their defect detection ability is 50%
better than the one observed in the human case
presented in their work, payback periods for incorpor-
ating an AVI system could be as low as 1 year.
A review of AVI research applied to the inspection
of wood boards showed that segmentation is often the
most time consuming part of the process, and one that
usually does not locate all defects properly (Pham
and Alcock 1998). Colour image segmentation algo-
rithms can be classified into one or more of the
following techniques (Cheng, Jiang, Sun and Wang
2001): histogram thresholding, feature space
clustering, region-based approaches, edge detection,
fuzzy approaches, neural networks, physics-based
approaches and hybrid techniques. The selection of
a colour space is application dependent. Brunner,
Maristany, Butler, VanLeeuwen and Funck (1992)
found that for wood images there is no advantage in
transforming the red, green and blue (RGB) colour
space into other colour spaces.
In Este
´vez, Perez and Goles (2003) a low-cost AVI
system for wood defect detection was introduced. The
system is composed by a standard colour video camera
connected to a PC Pentium IV; the lighting was a
mixture of two frontal halogen lights and ceiling
fluorescent lamps. The image segmentation module
operated using a histogram-based multiple threshold-
ing technique. The defect detection rate achieved was
94%. This high rate of defect detection was achieved
at the expense of increasing the rate of false-positives
(32%), i.e. dark grain lines segmented as defects. The
feature extraction module extracted 182 features from
the segmented defects and was used as inputs to a
neural classifier, which classified the defects into one of
the 11 defect categories. To improve the performance
achieved with individual classifiers, the combination of
three or five neural classifiers was carried out by simple
averaging of their outputs. All networks had one
hidden layer with an N
architecture, where N
is the number of inputs, N
is the number of hidden
units and N
is the number of outputs or categories.
Networks were trained by a second-order Back-
Propagation Quasi-Newton learning algorithm (BPQ)
(Saito and Nakano 1997), which uses an adaptive step-
length. The objective function minimised was the sum
of square errors plus a penalty term consisting of the
sum over all squared weights. The penalty term was
weighted by a regularisation factor. For each simula-
tion, the weights were randomly initialised. Each
variable was normalised in the range [0,1], and the
database was randomly split into training, validation
and test sets. Experimental results showed a
performance of 75% of correct classification in the
11 class wood defect problem.
Amongst the computational intelligence techni-
ques, the flexibility of fuzzy sets and the computational
efficiency of neural networks have caused a great
amount of interest in the combination of both
techniques. In the so-called neurofuzzy approaches,
Simpson (1993) introduced the fuzzy min–max (FMM)
clustering neural network, where clusters are repre-
sented as hyperboxes in the n-dimensional pattern
space. In Gabris and Bargiela (2000), the FMM was
extended and improved.
In this work, we present improvements to a low-
cost AVI system by using an image segmentation
method based on the FMM neural networks. The
method is called fuzzy min–max neural network for
image segmentation (FMMIS).
The FMMIS method
has been successfully applied to wood defect detection
in Ruz, Este
´vez and Perez (2005) achieving a defect
detection rate of 95% with a false positive rate of 6%
on 300 test wood board images with an average time
processing of 0.11 s per image, other applications
include face detection (Este
´vez, Flores and Perez
2005). Also, additional features are used in the
feature extraction module and the comparison of
Multi-Class Support Vectors Machines with MLP
neural networks in the classification module is
2. Image segmentation method
The image segmentation method for wood
defect detection consists of four stages, shown in
Figure 1. Each stage is described in the following
2.1. Seed selection process
To speed up the image segmentation process, the
FMMIS does not use all the pixels from the image
analysed. Instead, it only uses a few input pixels, called
seeds, to grow the hyperboxes. The seeds are auto-
matically determined (located) by an ad hoc procedure,
which is described in what follows. Considering the
great variability of colour of the wood boards, the seed
selection is based on adaptive thresholding. For each
board image, the mean colour intensity per channel,
, and the minimum colour intensity per channel, "
where t¼R,G,B, are measured. A cumulative histo-
gram per channel, H
, is defined as:
164 G.A. Ruz et al.
Downloaded By: [Ruz, Gonzalo A.] At: 15:49 16 February 2009
where nis the colour intensity level (n$255) and h
the histogram of the wood board image for channel t.
In addition, a colour intensity level is selected,
per channel, based on the cumulative histogram as
ðÞ ð2Þ
where $is a user-defined parameter, typically $$0:01.
To detect defects that are brighter than those detected
by using #t, an additional colour intensity is consid-
ered, and calculated as:
For each wood board image, the seeds are the pixels
belonging to the following intensity range:
½'^%tif #t<&
½' if #t(&t
where &tis a user-defined threshold for each channel,
that allows avoiding false positives when all the
defects on the wood board image are not too dark.
In Figure 1(a) the seeds are represented as white
2.2. Input patterns
The input patterns are the spatial coordinates of the
seeds, determined in Section 2.1 with each dimension
normalised in the range [0, 1]. Let Xbe a S)2 input
matrix (Figure 1(b)), where Sis the number of
seeds selected. The position of the h-th seed in
the image is represented by the vector Xh¼
ðÞ2I2, where the first coordinate indicates
the column and the second coordinate the row of the
2.3. Fuzzy min–max neural network for image
The FMMIS is built using hyperbox fuzzy sets. A
hyperbox defines a region in the n-dimensional
geometric space by pairs of min–max points for each
spatial coordinate of the image (rectangular boxes in
the case of two-dimensional images). Each hyperbox
fuzzy set has associated a membership function, which
describes the degree of membership (spatial proximity)
of a given pixel to a hyperbox in the [0,1] interval.
Let each hyperbox fuzzy set, B
, be defined by the
ordered set
where V
) is the min-point and W
is the max-point. The membership function for the
j-th hyperbox is 0 $bjðXh,Vj,WjÞ$1. Seeds contained
within a hyperbox have full membership value equal
to 1. The more distant the seeds are from the min–max
bounds of the hyperbox, the lower are their member-
ship values. The membership function defined in
Gabris and Bargiela (2000) is used here:
i¼1,2 %min %1*fðxhi *!ji,(
1*fð'ji *xhi,(
(( ð6Þ
where fis the ramp function defined as,
and (is the sensitivity parameter, that controls how
fast the membership value decreases when the seed is
farther from the min–max bounds of the hyperbox.
Figure 2 shows the two-dimensional membership
function of (6), with min-point V¼[0.4 0.4] and
Figure 1. Different stages of the image segmentation method: (a) seed selection process, (b) input patterns, (c) FMMIS and
(d) minimum bounding rectangles enclosing the defects, which correspond to the FMMIS output.
International Journal of Systems Science 165
Downloaded By: [Ruz, Gonzalo A.] At: 15:49 16 February 2009
max-point W¼[0.7 0.7], and (¼1. The membership
values ranging from 0 to 1 are represented by the gray-
scale ranging from black to white, respectively.
2.3.1. FMMIS learning algorithm
(1) Initialisation: V
and W
points are initially set
to 0. When a hyperbox is adjusted for the first
time using the input pattern Xh¼ðxh1,xh2Þ, the
min and the max points of the hyperbox are
made identically to this pattern,
(2) Hyperbox expansion: When an input pattern
(new seed pixel) is presented, the hyperbox with
the highest degree of membership is found and
expanded to enclose the input pattern. The
hyperbox expansion is accepted only if the
region contained by the expanded hyperbox is
similar in colour to the region enclosed by the
hyperbox before the expansion. A fuzzy colour
homogeneity criterion based on the standard Z
function of the Euclidean distance between the
mean colour intensities of the two hyperboxes
is defined. In this way, the colour similarity in
the RGB space between the hyperboxes before
and after the expansion is compared. A user-
defined parameter )2[0,1] is introduced to
control the required degree of colour homo-
geneity for expanding hyperboxes.
Formally, the following constraint must be
satisfied to expand a hyperbox,
where Zis a fuzzy membership function
(Cheng, Jiang and Wang 2002) defined as
and xis the Euclidean distance between the
mean colour intensities in the image region
covered by the two hyperboxes (before and after
expansion), measured in the RGB space. The
parameters of (10) are set to: a¼0, b¼L=2,
c¼Land L¼255 ffiffi
p. If the expansion criterion
for including the current seed is not satisfied, a
new hyperbox is created starting with that seed
as done in (8).
(3) Hyperbox overlap test: Let us assume that
the hyperbox B
was expanded in the previous
step. To test for overlapping, a dimension-
by-dimension comparison is performed
between B
and all the rest B
with k6¼j.
Overlap exists between B
and B
, if one of
the following four cases are met for dimen-
sions i¼1, 2:
Case 1: 'ji <'
ki <!
ji <!
Case 2: 'ki <'
ji <!
ki <!
Case 3: 'ji <'
ki $!ki <!
Case 4: 'ki <'
ji $!ji <!
(4) Hyperbox contraction: If overlap exists
between B
and B
, both hyperboxes begin to
contract until the overlap is eliminated. The
hyperbox contraction rules, which depend on
the four cases previously described, are as
Case 1: 'ji <'
ki <!
ji <!
ki ¼!new
ji ¼'old
ki þ!old
Case 2: 'ki <'
ji <!
ki <!
ji ¼!new
ki ¼'old
ji þ!old
Case 3: 'ji <'
ki $!ki <!
If !ki *'ji <!
ji *'ki then
ji ¼!old
ki ð13Þ
ji ¼'old
ki ð14Þ
00.2 0.4 0.6 0.8 1
Figure 2. Example of the two-dimensional membership
function associated to each hyperbox used by the FMMIS.
166 G.A. Ruz et al.
Downloaded By: [Ruz, Gonzalo A.] At: 15:49 16 February 2009
Case 4: 'ki <'
ji $!ji <!
by symmetry the same assignments as in
Case 3.
(5) Fine-tuning hyperbox expansion: After a
single pass through all the seeds, there is a
fine-tuning hyperbox expansion process, which
allows the hyperbox to grow, if necessary,
until the defect is completely enclosed. For
two-dimensional images, the hyperboxes are
rectangles defined by four line segments. A
horizontal line segment, hmin, and a vertical
line segment, vmin, pass through the min-
point of the hyperbox. Likewise, a horizontal
line segment hmax, and a vertical line
segment, vmax, pass through the max-point
of the hyperbox. For the hyperbox B
, and the
colour channel t, the following notation is
introduced: hmax
(r), vmax
(r), hmin
(r) and
(r) are the colour intensities of the r-th
pixel belonging to the line segments hmax,
vmax, hmin and vmin, respectively. Let u
a colour intensity threshold defined for
channels t¼R,G,B. A line segment (any of
the four) would be expanded if it contains
pixels darker than u
. The following conditions
for the while cycles should be satisfied for each
Case 1: while ðmaxrðhmaxjtðrÞÞ $ utÞ
j2þ1, update hmaxjt with !new
Case 2: while ðmaxrðvmaxjtðrÞÞ $ utÞ
j1þ1, update vmaxjt with !new
Case 3: while ðmaxrðhminjtðrÞÞ $ utÞ
j2*1, update hminjt with 'new
Case 4: while ðmaxrðvminjtðrÞÞ $ utÞ
j1*1, update vminjt with 'new
(6) Hyperbox merging: After finishing the
fine-tuning hyperbox expansion process
(if necessary), a final step is added in order to
merge hyperboxes belonging to the same defect.
The number of hyperboxes constructed per
defect depends on the value of )in (9). Let c
be the number of hyperboxes after the
fine-tuning hyperbox expansion process.
The centroid, CB
), of the hyperbox
is computed as:
cbji ¼'ji þ!ji
for i¼1, 2 and j¼1, ...,c. Let I
be the image
region contained within the limits of the
hyperbox B
. Let d
) be the Euclidean
distance in the colour space between the mean
intensities of the image regions I
and I
. The
fuzzy membership function !
, which mea-
sures the proximity in colour and space
between two hyperboxes, is defined as:
!c¼min Zd
where Zis the Z-function defined in (10), and
is the membership function defined in (6).
The merging criterion is to merge hyperboxes
whose proximity defined by (20) is greater than
a given threshold.
The FMMIS representation as a neural network
(Figure 1(c)) makes it possible to explore the paralle-
lism of the algorithm. Although the learning algorithm
is not necessarily neural (there is no biological principle
underlying the expansion–contraction process), the
execution of the network, once trained, fits in a
neural scheme. For this case, a three-layered neural
network was chosen to implement the FMMIS, as
shown in Figure 1(c). The input layer, F
, consists of
two processing elements (PE), one for each dimension
of the input pattern (seed) Xh¼ðxh1,xh2Þ. The second
layer, F
, consists of mPE (ideally one per each defect
in the wood board image). Finally, the third layer, F
does the merging of the nodes of the second layer
belonging to the same defect. There are two connec-
tions between each node of F
and each node of F
(Figure 1(c)); this is because the input can be crisp or
fuzzy (min–max value). Each node of F
in the three-
layered neural network represents a hyperbox fuzzy
set, where the connections from F
to F
are the min-
and max-points. The transference function of F
is the
membership function defined in (6). The connections
are adjusted using the learning algorithm described in
this subsection. The connections between the nodes
of F
and F
have binary values and they are stored
in the Umatrix, with value 1 if b
follows the
merging criterion giving origin to c
, and with value
0 otherwise.
2.4. Output of the FMMIS
The last stage is to draw the rectangle (ideally a
minimum bounding rectangle) on each defect using
International Journal of Systems Science 167
Downloaded By: [Ruz, Gonzalo A.] At: 15:49 16 February 2009
the min- and max-points of each hyperbox
formed by the FMMIS algorithm, as shown in
Figure 1(d).
3. Feature extraction
The feature extraction module of the previously
developed AVI system (Este
´vez et al. 2003) extracted
features from objects and windows of 64 by 64 pixels
centred in the object geometrical centre. The features
used in this work include: 7 object geometrical features
measured on the binarised gray image (e.g. area,
perimeter, average radius, aspect ratio, etc.); 96
object colour features (24 features measured in each
of the four channels, R, G, B and gray); and
46 window colour features (e.g. mean and variance of
window histograms, mean and variance at the edge of
3.1. Additional features
In addition to the 149 features mentioned above, new
features were added. These features are computed
using the co-occurrence matrix of the defects contained
in the rectangles found by the FMMIS; they receive the
name of co-occurrence features. We considered these
features because they are useful for separating classes
like clear wood and stain where the texture of the
image is important.
To compute the co-occurrence features, the gray
levels of the image contained in a square window are
quantised into Qlevels. Then, a two-dimensional
matrix p(i,j) is constructed (0 5i,j5Q). Point (i,j)
in the matrix represents the number of times that a
pixel with level iis followed, at a distance d([0 1],
[*1 1], [*1 0] and [*1*1]) and angle #(0+, 45+, 90+
and 135+), by a pixel with level j. A number of these
matrices can be computed for angles at intervals of 45+
and for each value of d. A more detailed description
on how to calculate the co-occurrence matrix can be
found in Gonzalez and Wood (2002), and Pham and
Alcock (2003).
A large number of features can be derived from
these matrices. In our work, we computed the
following features:
Contrast ¼X
Correlation ¼PQ
Energy ¼X
Homogeneity ¼X
In the extraction of the co-occurrence features we used
Q¼64. Also, for each of the four angles in the
co-occurrence matrix, the four measures were com-
puted. So at the end, 16 new features were added.
4. Classification
With the new segmentation module and the 16
additional features, the use of other classification
techniques was explored to see if any extra improve-
ments could be obtained. In addition to testing the
original MLP neural networks, the multi-class support
vector machines (SVM) were explored as well.
The multi-class SVM employed in this work were the
pairwise classification, one versus the rest, and the
binary decision tree.
The SVM pairwise classification was introduced
in Friedman (1996) and has become one of the most
popular Multi-Class SVM due to its good performance
in several classification tasks (Osuna, Freund and
Girosi 1997; Mayoraz and Alpaydin 1999). The basic
idea behind this classifier is to create a binary SVM for
each possible pair of classes, that is, for a Kclass
problem there are K(K*1)/2 binary classifiers. Then
when a new example is presented for classification,
each trained SVM is evaluated with this example and
a vote is assigned to the winning class. This method is
called the voting scheme, and the class label for
the example is the class that obtained the majority of
the votes.
The one versus the rest (Vapnik 1998)
classification scheme is built using Kbinary SVMs,
168 G.A. Ruz et al.
Downloaded By: [Ruz, Gonzalo A.] At: 15:49 16 February 2009
used to find a decision boundary that will separate a
certain class from the rest of the classes. This method
is faster than pairwise classification since only a few
SVMs are involved.
The SVM binary decision tree (Bennett 1999; Platt,
Cristianini and Shawe-Taylor 2000) has a binary tree
architecture, where the nodes are binary classifiers.
The nodes can contain multiple classes which are split
into two equally sizes subsets while moving along in
depth of the tree. This classifier is faster than the
previous two, since it uses less binary classifiers.
5. Methods
To test the overall performance of the AVI system
for wood defect detection, colour wood board
images (320 )240 pixels) drawn from the
University of Chile database (Este
´vez et al. 2003)
was processed and used for the FMMIS. All the
objects segmented by the FMMIS were manually
labelled (there can be more than one object per
wood board image) into one of the following 10
defect categories (see Ruz, Este
´vez and Perez 2005
for details): birdseye (be), pockets (po), wane (wa),
split (sp), stain (st), blue stain (bs), pith (pi), dead
knot (dk), live knot (lk) and hole (ho). Also, the
clear wood (cl) category was added, which corre-
sponds to non-defective areas that the FMMIS
segmented as a defective one. This error is then
hopefully detected in the classification stage. The
number of wood board images used was enough to
obtain 200 examples per each defect category. Then,
the samples were split into 1100 for training, 550 for
validation and 550 for testing.
Using the results obtained in Ruz et al. (2005)
and in Ruz and Este
´vez (2005), the parameters of
FMMIS were set to, (¼1 (sensitivity parameter),
)¼0.99 (degree of colour homogeneity used in hyper-
box expansion), u
¼195 (fine-tuning hyperbox
expansion parameter) and D¼0.95 (hyperbox merging
For the multi-class SVMs, a lineal kernel, a
polynomial kernel, and a Gaussian kernel were
explored using the training set. Out of the three, the
best classification performance on the validation set
was obtained by a third-order polynomial kernel.
So this kernel was used for each multi-class SVM
when testing the overall performance.
The MLP neural network was trained by a second-
order quasi-Newton learning algorithm called BPQ
(Saito and Nakano 1997). A weight decay penalty term
was added to the cost function with a regularisation
factor ", in order to avoid overfitting and improve the
generalisation performance.
The test set was used to evaluate the correct
classification percentage on the three multi-class
SVMs and the MLP neural network. Results without
and with the added texture features were taken into
account as well.
Also, in order to identify which are the most
informative and relevant features for solving the
classification problem, a feature selection method
based on mutual information called AMIFS (Tesmer
and Este
´vez 2004) was used to rank the 165 features
from the most informative to the least informative.
The AMIFS adaptively controls the trade-off between
eliminating irrelevance or redundancy, avoiding the
need of a user defined parameter, and can handle data
of mixed nature (discrete and continuous features).
6. Experimental results
For the SVMs classification methods, the correct
classification percentage was computed taking the
average value of five different test runs.
For the pairwise classification, results on the test
set are shown in Table 1. The average correct
classification performance was 91.39%. Notice that
for this type of SVM, 11(11 *1)/2 ¼55 binary SVMs
were needed to be trained. The confusion matrix of the
best solution (test run 4) is shown in Table 2. The
numbers in the confusion matrix are the classification
percentages where each row adds up to 100. We can
appreciate that there are eight classes with a classifica-
tion rate greater than 90%, amongst them, the clear
wood class increased from 77% (previous low-cost
AVI system) to 92%. This result shows that the central
problem of the previous system, which was the
confusion with the clear wood class, is overcome.
Currently, the principal confusions are stain with blue
stain and blue stain with clear wood. Nevertheless, the
classification rate of stain increased from 40% (Este
et al. 2003) to 66%.
Table 1. Pairwise classification performance without and
with co-occurrence features.
Test run
classification %
without co-occurrence
classification %
with co-occurrence
1 87.45 91.45
2 86.91 90.91
3 88.91 91.64
4 88.36 91.69
5 87.09 91.27
Average 87.74 ,0.86 91.39 ,0.32
International Journal of Systems Science 169
Downloaded By: [Ruz, Gonzalo A.] At: 15:49 16 February 2009
We can appreciate that the classifier still presents
some confusion between stain, blue stain and clear
Table 3 shows the results obtained using the one
versus the rest, reaching an average performance of
86.21%. Here, only 11 SVM were needed to be trained.
For the SVM binary decision tree, the optimal division
of all the classes into two groups was searched first.
This was done by testing different group divisions,
searching for the one that obtained best classification
performance. In total, 462 binary classifiers were tested
finding two optimal divisions that obtained the same
results. The two possible divisions are shown in
Table 4. Using any of the two configurations, the
average performance is 82.79% shown in Table 5.
Table 6 shows the results obtained by the MLP
neural networks, for different values of the regularisa-
tion parameter "using two different architectures, one
with inputs N
¼165, hidden units N
¼15 and outputs
¼11 and the other N
¼165, N
¼25 and N
Using 2000 training epochs the best performance was
found with "¼0.001 achieving 83.88%.
In general, the performance increases with the
incorporation of the co-occurrence features.
Table 1 shows a 4% increase in the average perfor-
mance when these features are considered.
The ranking of the features using AMIFS is shown
in Figure 3. The 165 features are listed left to right
starting from the most informative feature ‘116’ to
the less informative one ‘6’. From this result, it is
important to point out that all the new features
(16 co-occurrence features) appear in the top part of
Table 2. Confusion matrix when using the pairwise classification.
be po wa sp bs st pi dk lk ho cl
be 1 0 0 0 0 0 0 0 0 0 0
po 0 0.88 0.02 0.02 0 0 0.06 0 0.02 0 0
wa 0 0 1 0 0 0 0 0 0 0 0
sp 0.02 0 0 0.98 0 0 0 0 0 0 0
bs 0 0 0 0 0.84 0.02 0 0 0 0 0.14
st 0 0 0 0 0.18 0.66 0 0 0.08 0 0.08
pi 0 0.04 0 0 0 0 0.96 0 0 0 0
dk 0 0.06 0 0 0 0 0 0.92 0 0.02 0
lk 0 0.02 0 0 0 0.02 0 0.04 0.92 0 0
ho 0 0 0 0 0 0 0 0 0 1 0
cl 0 0 0 0 0.08 0 0 0 0 0 0.92
Table 4. Division of the 11 defect categories into two
optimal subsets.
Configuration Subset 1 Subset 2
1 Birdseye Pockets
Wane Pith
Split Dead knot
Blue stain Live knot
Stain Hole
Clear wood
2 Birdseye Pockets
Split Wane
Blue stain Pith
Stain Dead knot
Clear wood Live knot
Table 3. One vs. the rest performance without and with
co-occurrence features.
Test run
classification %
without co-occurrence
classification %
with co-occurrence
1 83.36 86.32
2 84.57 86.56
3 83.89 86.91
4 84.05 85.48
5 84.11 85.76
Average 83.99 ,0.44 86.21 ,0.58
Table 5. SVM binary decision tree performance without and
with co-occurrence features.
Test run
classification %
without co-occurrence
classification %
with co-occurrence
1 80.95 82.35
2 81.15 82.95
3 81.37 83.89
4 80.88 82.56
5 80.99 82.21
Average 81.07 ,0.19 82.79 ,0.67
170 G.A. Ruz et al.
Downloaded By: [Ruz, Gonzalo A.] At: 15:49 16 February 2009
the ranking (‘150–165’ with shaded colour) meaning
that they are very informative, thus, useful for this
classification task. Other relevant features include:
‘116’ brightest part of a histogram from a centred
window on the object (blue channel), ‘96’ brightest part
of a histogram from the object (blue channel), ‘98’
darkest part of a histogram from a centred window on
the object (red channel), ‘78’ darkest part of a
histogram from the object (red channel), ‘12’ colour
variance (red channel), ‘15’ colour variance (gray
channel), ‘13’ colour variance (green channel), ‘42’
variance in the minimum bounding box of the object
(red channel) and ‘4’ the aspect ratio (height/width).
7. Conclusions
In an AVI system for wood defect classification, each
stage plays a key role in the overall performance.
Nevertheless, the segmentation stage is critical since
the features used as inputs to the classifiers are
computed from the segmented defects. Advanced
classification techniques will not yield good results
if features are not obtained from proper segmented
defects. A high false positive detection rate (clear
wood) would produce erroneous classification of
non-defective areas as well.
The segmentation method called FMMIS is based
on the original FMM, but with a new learning
algorithm specially adapted for image segmentation
tasks. The FMMIS method combines clustering with
region-based techniques to obtain a substantially
different method than the original Simpson’s FMM.
The feature extraction module improved the
classification when the co-occurrence features were
added. The importance of these features, for this
particular classification problem, was also assessed by
the results of the ranking of the features using AMIFS.
The pairwise classification outperformed the MLP
neural network as well as the other SVM’s tested,
reaching 91.39% of correct classification. This is a
substantial improvement compared to the old system,
which reached 75% of correct classification.
Although FMMIS performs a coarse segmenta-
tion, enclosing the defects by rectangles (ideally the
MBR), results show that this level of segmentation is
enough to achieve good classification performances
and also provides correct cutting boundaries of
defective areas for the rough mill, since these are
done vertically or horizontally. Because a detailed
segmentation is not needed, one of the advantages of
FMMIS is that it performs the segmentation very
fast: in average each wood board image is segmented
in 0.11 s.
Future work should address how to reduce the
confusion between the stain, blue stain and clear wood
categories. One option could be to construct special
features for this purpose, another could be to train an
additional classifier destined to solve this specific
This work was supported in part by Conicyt-Chile, under
grant Fondecyt 1050751.
1. This article is an extended version of Ruz and Este
Figure 3. Ranking of the features obtained by AMIFS.
The most informative to the less informative features are
presented left to right, the shaded features are the
co-occurrence features.
Table 6. MLP neural network performance without and
with co-occurrence features for two different network
architectures and three values of the weight-decay regular-
isation parameter ".
% without
% with
165–15–11 0.1 78.24 79.86
165–15–11 0.01 80.65 83.09
165–15–11 0.001 81.92 83.88
165–25–11 0.1 78.55 79.09
165–25–11 0.01 80.07 81.75
165–25–11 0.001 80.93 82.77
International Journal of Systems Science 171
Downloaded By: [Ruz, Gonzalo A.] At: 15:49 16 February 2009
Abouelela, A., Abbas, H.M., Eldeeb, H., Wahdan, A.A., and
Nassar, S.M. (2005), ‘Automated Vision System for
Localizing Structural Defects in Textile Fabrics’, Pattern
Recognition Letters, 26, 1435–1443.
Bennett, K.P. (1999), ‘Combining Support Vector and
Mathematical Programming Methods for Classification’,
in Advances in Kernel Methods: Support Vector Learning,
eds. B. SchYo
¨lkopf, C.J.C. Burges, and A.J. Smola,
Cambridge, MA: MIT Press, pp. 307–326.
Brunner, C.C., Maristany, A.G., Butler, D.A., VanLeeuwen,
D., and Funck, J.W. (1992), ‘An Evaluation of Color
Spaces for Detecting Defects in Douglas-fir Veneer’,
Industrial Metrology, 2, 169–184.
Buehlmann, U., and Thomas, R.E. (2002), ‘Impact of Human
Error on Lumber Yield in Rough Mills’, Robotics and
Computer Integrated Manufacturing, 18, 197–203.
Campoy, P., Canaval, J., and Pen
˜a, D. (2005), ‘InsPulp-I!:
An On-line Visual Inspection System for the Pulp
Industry’, Computers in Industry, 56, 935–942.
Cheng, H.D., Jiang, X.H., Sun, Y., and Wang, J. (2001),
‘Color Image Segmentation: Advances and Prospects’,
Pattern Recognition, 34, 2259–2281.
Cheng, H.D., Jiang, X.H., and Wang, J. (2002), ‘Color Image
Segmentation Based on Homogram Thresholding and
Region Merging’, Pattern Recognition, 35, 373–393.
Derganc, J., Likar, B., Bernard, R., Tomazˇ evic
ˇ, D., and
ˇ, F. (2003), ‘Real-time Automated Visual Inspection
of Color Tablets in Pharmaceutical Blisters’, Real-time
Imaging, 9, 113–124.
Enke, D., and Dagli, C. (1997), ‘Automated Misplaced
Component Inspection for Printed Circuit Boards’,
Computers and Industrial Engineering, 33, 373–376.
´vez, P.A., Perez, C.A., and Goles, E. (2003), ‘Genetic
Input Selection to a Neural Classifier for Defect
Classification of Radiata Pine Boards’, Forest Products
Journal, 53, 87–94.
´vez, P.A., Flores, R.J., and Perez, C.A. (2005), ‘Color
Image Segmentation Using Fuzzy Min-max Neural
Networks’, in Proceedings of the International Joint
Conference on Neural Networks, Montreal, Canada,
pp. 3052–3057.
Friedman, J. (1996), ‘Another Approach to Polychotomous
Classification’. Stanford University Technical Report.
Available online at:
Gabris, B., and Bargiela, A. (2000), ‘General Fuzzy Min-max
Neural Networks for Clustering and Classification’,
IEEE Transactions on Neural Networks, 11, 769–783.
Gonzalez, R.C., and Woods, R.E. (2002), Digital Image
Processing (2nd ed.), New Jersey: Prentice Hall.
Malamas, E.N., Petrakis, E.G.M., Zervakis, M., Petit, L.,
and Legat, J. (2003), ‘A Survey on Industrial Vision
Systems, Applications and Tools’, Image and Vision
Computing, 21, 171–188.
Mayoraz, E., and Alpaydin, E. (1999), ‘Support Vector
Machines for Multiclass Classification’, in International
Workshop on Artificial Neural Networks (IWANN99), eds.
J. Mira, and J.V. Sa
´s, pp. 833–842.
Osuna, E., Freund, R., and Girosi, F. (1997), ‘Training
support vector machines: An application to face detection’,
in IEEE Conference on Computer Vision and Pattern
Recognition, pp. 130–136.
Pham, D.T., and Alcock, R.J. (2003), Smart Inspection
Systems, London: Academic Press.
Pham, D.T., and Alcock, R.J. (1998), ‘Automated Grading
and Defect Detection: A Review’, Forest Products Journal,
48, 34–42.
Platt, J., Cristianini, N., and Shawe-Taylor, J. (2000), ‘Large
Margin DAGs for Multiclass Classification’, in Advances
in Neural Information Processing Systems, eds. S.A. Solla,
T.K. Leen, and K.R. Myuller, Cambridge, MA: MIT
Press, pp. 547–553.
Ruz, G.A., and Este
´vez, P.A. (2005), ‘Image Segmentation
Using Fuzzy Min-max Neural Networks for Wood Defect
Detection’, in Intelligent Production Machines and Systems-
First I*PROMS Virtual Conference, eds. D.T Pham,
E.E. Eldukhri, and A.J. Soroka, pp. 183–188.
Ruz, G.A., Este
´vez, P.A., and Perez, C.A. (2005), ‘A
Neurofuzzy Color Image Segmentation Method for
Wood Surface Defect Detection’, Forest Products
Journal, 55, 52–58.
Saito, K., and Nakano, R. (1997), ‘Partial BFGS Update and
Efficient Step-length Calculation for Three-layer Neural
Networks’, Neural Computation, 9, 123–141.
Simpson, P.K. (1993), ‘Fuzzy Min-max Neural Networks.
Part 2. Clustering’, IEEE Transactions on Fuzzy Sets, 1,
Tesmer, M., and Este
´vez, P.A. (2004), ‘AMIFS: Adaptive
Feature Selection by using Mutual Information’, in
Proceedings of IEEE International Joint Conference on
Neural Networks (IJCNN’04), Budapest, Hungary,
pp. 303–308.
Vapnik, V. (1998), Statistical Learning Theory, New York:
Zhang, L., Dehghanib, A., Sua, Z., Kingb, T., Greenwooda,
B., and Levesleyb, M. (2005), ‘Real-time Automated Visual
Inspection System for Contaminant Removal from Wool’,
Real-time Imaging, 11, 257–269.
172 G.A. Ruz et al.
Downloaded By: [Ruz, Gonzalo A.] At: 15:49 16 February 2009
... In recent years, with the development of machine learning methods, computer intelligent identification methods based on machine learning have been widely applied in various fields to provide reliable predictions to improve production efficiency [3]. Memetic algorithm with compromise search (MACS) [1], robust Gaussian regression filters (RGRF) [7], Neural Networks [17], and feature-based wood defect machine learning methods, including artificial neural network, support vector machines (SVM), and other methods [18,20,25,32,34], are also used to detect wood defects and surface to solve the shortcomings of manual inspection. Wu et al. proposed a defect detection method based on affinity propagation clustering which effectively improves the clustering speed with an accuracy of 87.68% by extracting color matrix features, multi-scanning the image, and adjusting the sliding window automatically [32], but the time consumption of the improved affinity propagation algorithm is 3.09 s which was far from meeting real-time requirements. ...
... However, this method requires experienced technicians to carry out early identification of wood defect characteristics which is difficult to be applied in actual working conditions. Fuzzy min-max neural networks (FMMIS) were presented by Ruz et al. [25] to improve the detection accuracy to 91% by ideally setting the defects in the wood image as bounded rectangles. One of the advantages of FMMIS is that it performs the segmentation very fast: the wood board image is segmented in 0.11 s for average, this performance still has room for further improvement. ...
... In addition, the detection accuracy has been further improved over the models based on the Affinity Propagation clustering method [32] and the neural network method [25] used in the previous literature. Then, we compared different algorithms of per image average recognition time. ...
Full-text available
Wood processing is one of the most widely used in agriculture and industry. Low precision and high time delay of machine learning in wood defect detection are currently the main factors restricting the production efficiency and product quality of the wood processing industry. An SPP-improved deep learning method was proposed to detect wood defects based on the basic framework of the YOLO V3 network to improve accuracy and real-time performance. The extended dataset was firstly established by image data enhancement and preprocessing based on the limited samples of the wood defect dataset. Anchor box scale re-clustering of the wood defect dataset was carried out according to the defect features. The spatial pyramid pooling (SPP) network was applied to improve the feature pyramid (FP) network in YOLO V3. The validity and real-time performance of the proposed algorithm were verified by a randomly selected test set. The results show that the overall detection accuracy rate on the wood defect test dataset reaches 93.23% while the detection time for each image is within 13 ms.
... In recent years, machine vision-based image inspection technology has developed rapidly. The traditional machine vision method of recognizing board defects is to identify board defects after steps such as greyscale transformation, smoothing filtering, threshold segmentation, edge detection, and contour extraction [15][16][17]. The accuracy of this recognition method is low, stability is poor, and the detection speed does not meet the real-time requirements. ...
... For defects such as image noise and low contrast, Qi et al [28] proposed a BP neural network for fast and accurate identification of defects inside the template, analysed the results of different network structures and network parameters affecting the classification of wood defects, and proposed optimal network parameters for identifying the types of wood panel defects. In addition to this, Ruz et al [15] used a fuzzy minmax neural network for image segmentation FMMIS to generate a minimum bounding rectangle for each defect present in the wood panel image with a correct classification rate of 91%. For knot and crack defects in wood panels, in 2015 Mohamad et al [29] proposed a dictionary based on SURF and LBP features and a bag-of-words approach for classification using SVM and showed through experimental results on two different datasets that the combination of the two features outperformed the single feature with a detection accuracy of 91.5%. ...
In recent years, deep learning has made significant progress in wood panel defect detection. However, there are still challenges such as low detection , slow detection speed, and difficulties in deploying embedded devices on wood panel surfaces. To overcome these issues, we propose a lightweight wood panel defect detection method called YOLOv5-LW, which incorporates attention mechanisms and a feature fusion network.Firstly, to enhance the detection capability of acceptable defects, we introduce the Multi-scale Bi-directional Feature Pyramid Network (MBiFPN) as a feature fusion network. The MBiFPN reduces feature loss, enriches local and detailed features, and improves the model's detection capability for acceptable defects.Secondly, to achieve a lightweight design, we reconstruct the ShuffleNetv2 network model as the backbone network. This reconstruction reduces the number of parameters and computational requirements while maintaining performance. We also introduce the Stem Block and Spatial Pyramid Pooling Fast (SPPF) models to compensate for any accuracy loss resulting from the lightweight design, ensuring the model's detection capabilities remain intact while being computationally efficient.Thirdly, we enhance the backbone network by incorporating Efficient Channel Attention (ECA), which improves the network's focus on key information relevant to defect detection. By attending to essential features, the model becomes more proficient in accurately identifying and localizing defects.We validate the proposed method using a self-developed wood panel defect dataset.The experimental results demonstrate the effectiveness of the improved YOLOv5-LW method. Compared to the original model, our approach achieves a 92.8\% accuracy rate, reduces the number of parameters by 27.78\%, compresses computational volume by 41.25\%, improves detection inference speed by 10.16\%
... In recent years, many scholars pay their attention to related algorithms on timber classification or defect detection. Prof Gonzalo [4] used a fuzzy min-max neural network for image segmentation. Then grade classification is performed based on a multilayer perceptron neural network. ...
Full-text available
The optimization of timber classification by grades and defect detection plays an important role in the production of timbers. Traditionally, a timber is manual cut by a worker according to his experience. Defect detection and classification of a timber are with great subjectivity. Meanwhile, the action is not safe enough. In this case, an automatic optimizing cross-cut saw to finish these tasks of timber classification by grades and defect detection is built significantly. Related algorithm and detailed procedure for optimizing cross-cut saw are proposed in this paper. Additionally, a vision system is used to capture images of a timber. Captured images are analyzed and processed. First, defects in these images are detected. Then the serviceable part (defect-free) of a timber can be determined. Based on the pretrained network, the timber can be classified. As the homography matrix has been known, the physical position can be confirmed. In our proposed system, the cutting list is transmitted from the industrial control computer to a motion control system, then the timber can be cut according to the cutting list automatically. In this paper, related algorithms and detailed procedure are given. Moreover, a new optimizing cross-cut saw is built. Experiments show that the processing time for each image is about 0.026s and the minimum mean average precision is 94.15%. In this case, it can make the optimizing cross-cut saw efficient, laborsaving and safe. Furthermore, related algorithms are suitable to improve a traditional automatic optimizing cross-cut saw.
... Analysis of the literature indicates that vision methods are widely used to control product parameters. The development of vision-based quality control methods has been progressing for many years [2]. Most of them are based on analysing two-dimensional images recorded using various methods. ...
Full-text available
This article presents a vision method of identifying and measuring wood surface parameters to detect defects resulting from errors occurring during machining. The paper presents the method of recording a three–dimensional image of the wood surface using the laser triangulation method. It discusses parameters related to imaging resolution and the impact of vision system configuration parameters on the measurement resolution and image acquisition time. For the recorded image, proposed algorithms detect defects like wade and bark at the board edges. Algorithms for measuring characteristic parameters describing the surface of the wood are presented. Validation tests performed using the prepared system in industrial conditions are provided and discussed. The proposed solution makes it possible to detect board defects in flow mode on belt conveyors operating at a speed of up to 1000 mm/s.
... The Bayesian regularization and backpropagation training methods outperformed the other competitors with an accuracy of 98.2%. The above discussion on wood defect identification based on various types of classifiers is summarized in Table 3. Decision tree [17], [59], [72] Random forest [32], [72], [74] Naïve Bayes [59], [71], [75], [76] SVM [16], [17], [24], [38], [45], [76]- [78] Neural network [8], [14]- [16], [29], [40], [46], [47], [49], [59], [75]- [77], [79]- [83] Lazy learner others k-NN [17], [27], [59], [72], [78] Particle swarm optimization [17], [27], [59], [72], [78] Genetic algorithms [73] Bees algorithms [84] ...
Full-text available
Timber quality control is undoubtedly a very laborious process in the secondary wood industry. Manual inspections by operators are prone to human error, thereby resulting in poor timber quality inspections and low production volumes. The automation of this process using an automated vision inspection (AVI) system integrated with artificial intelligence appears to be the most plausible approach due to its ease of use and minimal operating costs. This paper provides an overview of previous works on the automated inspection of timber surface defects as well as various machine learning and deep learning approaches that have been implemented for the identification of timber defects. Contemporary algorithms and techniques used in both machine learning and deep learning are discussed and outlined in this review paper. Furthermore, the paper also highlighted the possible limitation of employing both approaches in the identification of the timber defect along with several future directions that may be further explored.
... This drawback is considered and partly solved by Gabrys and Bargiela [28], Nandedkar and Biswas [38], Nandedkar and Biswas [75], Wang et al. [30]. [91,133] Biometrics Character recognition [92] Emotion recognition [128][129][130][131][132] Face detection [106] Hand gesture recognition [105] Human action recognition [66] Human motion recognition [134,135] Numeral character recognition [89,90,137] Sign language recognition [93][94][95] Signature recognition [96,97,136] Speaker/speech recognition [132] Wood defect detection, Face detection [155] Cybersecurity Attack intention recognition [50,51,60] Intrusion detection [102,168] General purpose Colour image segmentation [150,151] Colour video segmentation [154] Image retrieval [169] Object recognition [152,153] Shadow detection and removal [170] Wood defect detection [28,46,103,124,125,171,172] Industrial application Fault detection in a water system [45] Fault detection in cooling system [39] Fault detection in energy systems [173] Fault detection of ball bearing [99,101,[120][121][122][123][174][175][176][177] Fault detection of induction motors [43,70,126] Fault detection of oil pipeline [59] Motor fault detection [67,127] Pipeline defect detection [68,69] Power quality monitoring [144] Medical application Acute Coronary Syndrome [146] Brain glioma [139,140] Cervical cancer [141] Classification of chronic leukaemia [31] Classifying breast cancer patients [147,148] Fall detection [178] Genetics [48,145] Heart diseases [142,143] Liver disease [138] Lung cancer [178] Lung nodules detection [179] Medical diagnosis [180] Patient admission prediction [181] Signal recognition of brain [47] Stroke [149] Wheelchair control [160] Other Agricultural circular economy region division [165] Autonomous vehicle navigation [166] Fault detection in rail vehicle systems [164] Classification of music by composer [182] Clustering of radar pulses [163] Clustering of the satellite infrared images [71] Financial problems [162] Flood forecasting [167] Imputation of missing values in surveys of voting intention polls [160] Special education of learning disabled learners [156] Robotics Fault detection [157] Human-robot affective relationship [158] Human-robot affective relationship ...
Full-text available
The amount of digital data in the universe is growing at an exponential rate with the rapid development of digital information, and this reveals new machine learning methods. Learning algorithms using hyperboxes are a subsection of machine learning methods. Fuzzy min–max neural network (FMNN) are one of the most common and advanced methods using hyperboxes. FMNN is a special type of NeuroFuzzy system that combines the artificial neural network and fuzzy set into a common framework. This paper conducts an extensive bibliometric and network analysis of FMNN literature. Two hundred and sixty-two publications are analysed from the period of 1992–2022. Several analyses are realized in order to identify trends, challenges and key points in a more scientific and objective way that affect the development of knowledge in the FMNN domain. It can be seen from bibliometric analysis that there is rapid development in the last 10 years. Social network analysis results show that Chee Peng Lim is the most active author in the network. Besides, the modifications of FMNN are generally developed for classification. However, there are still potential future research opportunities for clustering.
... Currently, manufacturers of specialized machines are forced to equip the machines with control systems that carry out selected inspection tasks directly at the production stations. For example, studies are carried out to optimize the cutting of boards in sawmills, to control the surface of the wood in the furniture industry and monitoring of the absence of structural defects in boards as a construction material [3][4][5][6][7]. In addition, quality control of the finished products just before shipment to the customer is also realized to improve the companies' financial performance [8]. ...
Full-text available
Wood used in production processes can be infected by various fungi growing on its surface. The presence of fungi on the wood surface results from the method of storage, handling and transport of the wood. However, the presence of fungi on wood carries a high risk to the health of production operators and users. At the same time, it has a negative impact on the quality and durability of manufactured products. Because of the risks indicated, an attempt was made to develop an industrial, automated system for detecting fungal infections. This paper presents a vision method for detecting fungal infections on the wood surface. A description of the vision system using the laser triangulation method (LTM) to build a three-dimensional surface image is shown. The paper consists of an analysis of the imaging resolution and a description of the concept of using laser illuminator power selection for identifying fungal-infested surfaces. Imaging results for the selected wavelength of electromagnetic radiation are presented. Measurements and parameters describing the identified areas are shown. It was found that it is possible to choose imaging method parameters and laser illumination power allowing identification under industrial conditions of a fungus-infected region on a wood surface while using the image to determine product measurement parameters.
... Early studies used color and shape-based approaches to detect wood knots (Alapuranen and Westman 1992;Lampinen et al. 1994;Kauppinen and Silvén 1996), and after which the knot classification was treated as a matter of texture classification (Kamal et al. 2017;Mahram et al. 2012;Xie and Wang 2015). Gray-level co-occurrence matrix (GLCM)-based Haralick texture features (hereinafter referred to as GLCM features) was preferred in studies for automated wood defect classification, and promising results were produced from models trained with GLCM features (Kamal et al. 2017;Qayyum et al. 2016;Ruz et al. 2009;Xie and Wang 2015). Mahram et al. (2012) reported that models trained with multi-feature sets combined with Haralick and local binary pattern (LBP) features achieved better performance than those trained with each single feature set. ...
Full-text available
This paper describes feature-based techniques for wood knot classification. For automated classification of macroscopic wood knot images, models were established using artificial neural networks with texture and local feature descriptors, and the performances of feature extraction algorithms were compared. Classification models trained with texture descriptors, gray-level co-occurrence matrix and local binary pattern, achieved better performance than those trained with local feature descriptors, scale-invariant feature transform and dense scale-invariant feature transform. Hence, it was confirmed that wood knot classification was more appropriate for texture classification rather than an approach based on morphological classification. The gray-level co-occurrence matrix produced the highest F1 score despite representing images with relatively low-dimensional feature vectors. The scale-invariant feature transform algorithm could not detect a sufficient number of features from the knot images; hence, the histogram of oriented gradients and dense scale-invariant feature transform algorithms that describe the entire image were better for wood knot classification. The artificial neural network model provided better classification performance than the support vector machine and k -nearest neighbor models, which suggests the suitability of the nonlinear classification model for wood knot classification.
Deep learning has achieved certain results in the field of wood surface defect detection. To address the problems of low accuracy of the detection results of surface defects on boards, slow detection speed and large number of model parameters, this article take advantage of computer vision to improve the feature fusion module of YOLOX target detection algorithm, by adding efficient channel attention (ECA) mechanism, adaptive spatial feature fusion mechanism (ASFF) and improve the confidence loss and localization loss functions as Focal loss and Efficient Intersection over Union (EIoU) loss, to enhance the feature extraction ability and detection accuracy of the algorithm. Considering the depth and width of the model, the depth-separable convolution and optional multi-version algorithm are used to reduce the model parameters and computational effort to seek the optimal model. Experiments show that the improved model detects four types of defects in rubber timber with a considerable improvement and has significant advantages over other target detection algorithms.
Full-text available
A crucial step in developing automated visual inspection systems for wood boards is image segmentation, which aims to achieve a high defect detection rate with a low false positive rate (clear wood areas identified as defect areas). In this study, a neurofuzzy color image segmentation method for wood surface defect detection is proposed. The method is called fuzzy min-max neural network for image segmentation (FMMIS). The FMMIS method grows boxes from a set of pixels called seeds, to find the minimum bounded rectangle (MBR) for each defect present in the wood board image. An automatic method to select seeds from defective regions as starting points to FMMIS is also presented. The FMMIS method was applied to a set of 900 images of radiata pine boards, which included samples from the following 10 categories of defects: birdseye and freckle, bark and pitch pockets, wane, splits, blue stain, stain, pith, dead knots, live knots, and holes. The FMMIS achieved a defect detection rate of 95 percent on the test set, with only 6 percent of false positives. To measure the quality of segmentation, the area recognition rate (ARR) criterion was computed using as a reference the manually placed MBR for each defect. The ARR achieved 94.4 percent on the test set. Also a relative index was used to compare the quality of segmentation between FMMIS and the segmentation module of a previously developed system, based on histogram thresholding. The results show that FMMIS allows us to obtain significant improvements compared with previous work.
Smart Inspection Systems - Techniques and Applications of Intelligent Vision will enable engineers to understand the various stages of automated visual inspection (AVI) and how artificial intelligence can be incorporated into each stage to create "smart" inspection systems. The book contains many examples that illustrate and explain the application of conventional and artificial intelligence techniques in AVI. The text covers the whole AVI process, from illumination, image enhancement, segmentation and feature extraction, through to classification, and includes case studies of implemented AVI systems as well as reviews of commercially available inspection systems. Each chapter concludes with exercises. This book will be of interest to users and developers of commercial industrial inspection systems as well as researchers in the fields of machine vision, artificial intelligence and advanced manufacturing engineering.
In a plant that manufactures wood products, such as lumber, dimension stock, or veneer sheets, inspection of the products is a necessary part of the production process. At present, most inspection tasks are carried out manually. However, because of the high speed with which it is necessary to perform the operation and the stress involved, attempts have been made to automate this grading process. This paper surveys research in the field of automatic inspection of wood, particularly focusing on computer vision techniques. The methods are put into an Automated Visual Inspection framework, which is subdivided into commonly used modules for image acquisition, image enhancement, image subdivision, feature extraction, and classification.
A genetic algorithm was used to determine an appropriate set of features for automatic defect classification of radiata pine boards. The study was performed using a low-cost machine vision system composed of a color video camera, a frame grabber, and a micro- computer. The following 10 defect categories were considered, plus clear wood: birds eye & freckle, bark & pitch pockets, wane, split, blue stain, stain, pith, dead knot, live knot, and hole. A database was built containing color images of 2,958 board faces. A total of 16,800 feature vectors were extracted from these images, and partitioned into training, validation, and test sets. Each vector was com- posed of 182 features measured in the segmented objects and in windows around the objects. By using feature selection algorithms, 64 out of 182 original features were selected and used as inputs to a multilayer perceptron neural network classifier, without reducing the classification performance. Using the set of features evolved by a genetic algorithm, the best off-line performance obtained was 74.5 percent of correct classifications on the test set. The classification performance on a reduced database with 7 defect categories reached 87.8 percent. An online system evaluation yielded 80 percent of correct classifications with 10 defect categories plus clear wood. The study shows that the genetic selection of features allows us to identify the most relevant features for complex classification problems, such as wood defect classification, where the best features are unknown.
Color information can be valuable for detecting and classifying surface defects inwood, but its usefulness depends on the color data's format and the analysis technique used. This study investigates five color transforms that convert the National Television Standards Committee red, green, and blue (RGB) primary color space into other potentially more useful spaces. A quadratic classifier was used to evaluate the relative utility of the different color spaces in separating defects from clear wood. Images of Douglas-fir veneer with encased knots, intergrown knots, and pitch streaks were converted to the various color spaces and then analyzed. The results show, for the conditions in this study, that a two-dimensional feature space is sufficient for classification and that there are no practically important differences in performance among the different color spaces. Thus, for images of Douglas-fir veneer, it appears that there is no advantage in mathematically transforming the original RGB data into another color space.
Quality control is one of the basic issues in textile industry. Analysis of texture content in digital images plays an important role in the automated visual inspection of textile images to detect their defects. In this paper, a system for automated visual inspection of textiles is discussed. A detailed system configuration is presented and a fault detection algorithm is proposed. Industrial vision systems must operate in real-time, produce a low false alarm rate and be flexible to accommodate variations in inspection sites. This was the rationale behind developing a detection algorithm which employs simple statistical features (mean, variance, median). The intent was to utilize such features to make the calculations simple and fast for the system to be suitable for real-time applications. The performance of the system was evaluated on plain fabrics with different types of textile flaws. The results indicate that the system can detect flaws which vary drastically in physical dimension and nature with a very low false detection rate.
In the textile industry, scoured wool contains different types of foreign materials (contaminants) that need to be separated out before it goes into further processing, so that the textile machines are protected from damage and the quality of the final woollen products is ensured. This paper presents an automated visual inspection (AVI) system for detecting and sorting contaminants from wool in real time. The techniques were first developed in the lab and subsequently applied to a large-scale factory system. The combinative use of image processing algorithms in RGB and HSV colour spaces can segment 96% of contaminant types (minimum size around 4 cm long and 5 mm in diameter) in real-time on the lab test rig. One of the most important aspects of the system is to use the non-linear colour space transformation and merge the threshold algorithm in HSV colour space into the image processing algorithms in RGB colour space to enhance the contaminant identification in real time. The real-time capability of the system is also analysed in detail. The experimental results demonstrate that the factory AVI system could identify and remove the contaminants at a camera speed of around 800 lines/s and the conveyor speed of 20 m/min in real time.
Image segmentation is very essential and critical to image processing and pattern recognition. This survey provides a summary of color image segmentation techniques available now. Basically, color segmentation approaches are based on monochrome segmentation approaches operating in different color spaces. Therefore, we first discuss the major segmentation approaches for segmenting monochrome images: histogram thresholding, characteristic feature clustering, edge detection, region-based methods, fuzzy techniques, neural networks, etc.; then review some major color representation methods and their advantages/disadvantages; finally summarize the color image segmentation techniques using different color representations. The usage of color models for image segmentation is also discussed. Some novel approaches such as fuzzy method and physics-based method are investigated as well.