Conference PaperPDF Available

Texture classifiers generated by genetic programming

Authors:

Abstract

We investigate the behaviour of image texture classifiers generated by genetic programming. We propose techniques to understand how classifiers capture textural characteristics and for discussing the effectiveness of different classifiers. Our results show that regularities of patterns can be detected by the genetic programming method without predefined knowledge
0-7803-7282-4/02/$10.00 ©2002 IEEE
0-7803-7282-4/02/$10.00 ©2002 IEEE
0-7803-7282-4/02/$10.00 ©2002 IEEE
0-7803-7282-4/02/$10.00 ©2002 IEEE
0-7803-7282-4/02/$10.00 ©2002 IEEE
0-7803-7282-4/02/$10.00 ©2002 IEEE
... GP is a novel method to tackle a broad range of problems due to its flexibility and fluency of computer program representation as well as the strong proficiencies of its evolutionary search. GP applications have shown a trend of success in recent years [18][19][20][21][22][23]. The main advantage of GP algorithm is that it performs a global search for a model allowing evaluation of that model as a whole in the fitness function, without focusing on the impact of each possible condition. ...
... The model is created by determining a subset of data as training part by which an algorithm is trained to label the classes. Then, the model is applied on a different dataset, called test set to predict the class of each member of the dataset using the model learned from training [19]. In the most cases, the problem uses supervised training in which a portion of dataset labeled with the type of the class it belongs to is provided to the system. ...
Article
Full-text available
Genetic programming (GP) is an artificial intelligence technique that benefits from evolutionary computations allowing computers to solve problems automatically. In this paper, we present an optimized genetic-programming-based classifier that directly solves the multi-class classification problems in data mining and image analysis. A new fitness function is proposed for multiclass classification and brain tumor detection, which is validated by 10-fold cross validation. Instead of defining static thresholds as boundaries to differentiate between multiple labels, our work presents a method of classification where a GP system learns the relationships among experiential data and models them mathematically during the evolutionary process. We propose an optimized GP classifier based on a combination of pruning subtrees and a new fitness function. An orthogonal least squares algorithm is also applied in the training phase to create a robust GP classifier. Our approach has been assessed on six multiclass datasets and on a magnetic resonance imaging (MRI) brain image for tumor detection. The results of data classification for Iris, Wine, Glass, Pima, BUPA Liver and Balance Scale datasets are compared with existing algorithms. The high accuracy of brain tumor classification provided by our GP classifier confirms the strong ability of the developed technique for complicated classification problems. We compared our approach in terms of speed with previous GP algorithms as well. The analyzed results illustrate that the developed classifier produces a productive and rapid method for classification tasks that outperforms the previous methods for more challenging multiclass classification problems.
... GP is a novel method to tackle a broad range of problems due to the fact that it utilizes the flexibility and fluency of computer program representation as well as the strong proficiencies of evolutionary search. GP applications have showed a trend of success in recent years [14][15][16][17][18]. ...
... The modeled data is created by determining a set of data as training part by which the classes are labeled in order to teach the algorithm. Then, the model is applied on a different dataset called test set in order to predict the class of each member of the dataset using the model learned from training part [15]. In the most cases, the problem uses ...
Conference Paper
Full-text available
This paper¹ proposes an approach which uses optimized genetic programming (GP) with a new fitness function for multiclass dataset classification. In place of defining static thresholds as boundaries to differentiate between multiple classes, our work presents a method of classification where a GP system learns the relationships among experiential data and models them mathematically during the evolutionary process. We propose an optimized GP classifier based on a combination of pruning subtrees and a new fitness function. An orthogonal least squares algorithm is also applied in the training phase to create a robust GP classifier. Our approach has been assessed on four multiclass datasets and compared against three existing methods. The analyzed results illustrate that the developed classifier produces a productive and rapid method for classification tasks that outperforms the previous methods for more challenging multiclass classification problems.
... In terms of the number of classes in object classification, there are two categories: binary classification problems, where there are only two classes of objects to be classified, and multi-class classification problems, where more than two classes of images are involved. While GP has been widely applied to binary classification problems [38,44,10,45], it has also been applied to multi-class classification problems [46,22,47,16,15,48]. ...
Article
Full-text available
This paper describes two innovations that improve the efficiency and effectiveness of a genetic programming approach to object detection problems. The approach uses genetic programming to construct object detection programs that are applied, in a moving window fashion, to the large images to locate the objects of interest. The first innovation is to break the GP search into two phases with the first phase applied to a selected subset of the training data, and a simplified fitness function. The second phase is initialised with the programs from the first phase, and uses the full set of training data with a complete fitness function to construct the final detection programs. The second innovation is to add a program size component to the fitness function. This approach is examined and compared with a neural network approach on three object detection problems of increasing difficulty. The results suggest that the innovations increase both the effectiveness and the efficiency of the genetic programming search, and also that the genetic programming approach outperforms a neural network approach for the most difficult data set in terms of the object detection accuracy.
... Wavelets have been employed in various work including Chen and Lu (2007) and Padole and Athaide (2013). Texture features constructed from pixel grey levels were used in Song et al. (2002) to discriminate simple texture images. Cartesian GP was used to implement Transform based Evolvable Features (TEFs) in Kowaliw et al. (2009), which were used to evolve image transformations: an approach which improved classification of Muscular Dystrophy in cell nuclei by 38 % over previous methods . ...
Chapter
Full-text available
This chapter describes a general approach for image classification using Genetic Programming (GP) and demonstrates this approach through the application of GP to the task of stage 1 cancer detection in digital mammograms. We detail an automated work-flow that begins with image processing and culminates in the evolution of classification models which identify suspicious segments of mammograms. Early detection of breast cancer is directly correlated with survival of the disease and mammography has been shown to be an effective tool for early detection, which is why many countries have introduced national screening programs. However, this presents challenges, as such programs involve screening a large number of women and thus require more trained radiologists at a time when there is a shortage of these professionals in many countries.Also, as mammograms are difficult to read and radiologists typically only have a few minutes allocated to each image, screening programs tend to be conservative—involving many callbacks which increase both the workload of the radiologists and the stress and worry of patients.Fortunately, the relatively recent increase in the availability of mammograms in digital form means that it is now much more feasible to develop automated systems for analysing mammograms. Such systems, if successful could provide a very valuable second reader function.We present a work-flow that begins by processing digital mammograms to segment them into smaller sub-images and to extract features which describe textural aspects of the breast. The most salient of these features are then used in a GP system which generates classifiers capable of identifying which particular segments may have suspicious areas requiring further investigation. An important objective of this work is to evolve classifiers which detect as many cancers as possible but which are not overly conservative. The classifiers give results of 100 % sensitivity and a false positive per image rating of just 0.33, which is better than prior work. Not only this, but our system can use GP as part of a feedback loop, to both select and help generate further features.
... Zhang, Ciesielski and Andreae [8] describe the evolution of one-step classifiers for finding different classes of small objects in large images directly from pixel values. Bhanu and Lin [9] describe a similar approach to object detection, but the inputs [10] describe the evolution of one-step classifiers for texture classification by genetic programming. The evolved classifiers achieved good performance on classification problems involving Brodatz textures. ...
Conference Paper
Our primary motivation in this paper is to determine whether evolved texture feature extraction programs are competitive with human derived programs for a difficult real world texture classification problem. The problem involves distinguishing images of three classes of bulk malt. There are subtle differences between the three classes. We have used a number of human derived methods, Haralick, Gabor, Haar, histogram and Galloway, to get feature vectors for the malt problem. We have also used a number of feature extraction programs that were evolved from thirteen Brodatz textures. We performed classification with a 1 nearest neighbour classifier. The evolved features gave an accuracy of 67% which is considerably better than the 53% achieved with the Haar features, but not as good as the 77% achieved with the Galloway features. Analysis of the evolved features suggested that they are capturing some texture regularities not captured by the human derived methods. We conclude that the evolved features are competitive with the human derived features and can provide enhanced accuracy when used in conjunction with human derived features.
Article
Full-text available
Epilepsy, sometimes called seizure disorder, is a neurological condition that justifies itself as a susceptibility to seizures. A seizure is a sudden burst of rhythmic discharges of electrical activity in the brain that causes an alteration in behaviour, sensation, or consciousness. It is essential to have a method for automatic detection of seizures, as these seizures are arbitrary and unpredictable. A profound study of the electroencephalogram (EEG) recordings is required for the accurate detection of these epileptic seizures. In this study, an Innovative Genetic Programming framework is proposed for classification of EEG signals into seizure and nonseizure. An empirical mode decomposition technique is used for the feature extraction followed by genetic programming for the classification. Moreover, a method for intron deletion, hybrid crossover, and mutation operation is proposed, which are responsible for the increase in classification accuracy and a decrease in time complexity. This suggests that the Innovative Genetic Programming classifier has a potential for accurately predicting the seizures in an EEG signal and hints on the possibility of building a real‐time seizure detection system.
Article
In this paper, we present a new method for classification of electroencephalogram (EEG) signals using Genetic Programming (GP). The Empirical Mode Decomposition (EMD) is used to extract the features of EEG signals which served as an input for the GP. In this paper, new constructive crossover and mutation operations are also produced to improve GP. In these constructive crossover and mutation operators hill climbing search is integrated to remove the destructive nature of these operators. To improve GP, we apply constructive crossover on all the individuals which remain after reproduction. A new concept of selecting the global prime off-springs of the generation is also proposed. The constructive mutation approach is applied to poor individuals who are left after selecting globally prime off-springs. Improvement of the method is measured against classification accuracy, training time and the number of generations for EEG signal classification. As we show in the results section, the classification accuracy can be estimated to be 98.69% on the test cases, which is better than classification accuracy of Liang and coworkers method which was published in 2010.
Conference Paper
Full-text available
We describe an approach to the use of genetic programming for object detection problems in which the locations of small objects of multiple classes in large pictures must be found. The evolved programs use a feature set computed from a square input field large enough to contain each of objects of interest and are applied, in moving window fashion, over the large pictures in order to locate the objects of interest. The fitness function is based on the detection rate and the false alarm rate. We have tested the method on three object detection problems of increasing difficulty with four different classes of interest. On pictures of easy and medium difficulty all objects are detected with no false alarms. On difficult pictures there are still significant numbers of errors, however the results are considerably better than those of a neural network based program for the same problems.
Article
Full-text available
Texture is one of the important characteristics used in identifying objects or regions of interest in an image, whether the image be a photomicrograph, an aerial photograph, or a satellite image. This paper describes some easily computable textural features based on gray-tone spatial dependancies, and illustrates their application in category-identification tasks of three different kinds of image data: photomicrographs of five kinds of sandstones, 1:20 000 panchromatic aerial photographs of eight land-use categories, and Earth Resources Technology Satellite (ERTS) multispecial imagery containing seven land-use categories. We use two kinds of decision rules: one for which the decision regions are convex polyhedra (a piecewise linear decision rule), and one for which the decision regions are rectangular parallelpipeds (a min-max decision rule). In each experiment the data set was divided into two parts, a training set and a test set. Test set identification accuracy is 89 percent for the photomicrographs, 82 percent for the aerial photographic imagery, and 83 percent for the satellite imagery. These results indicate that the easily computable textural features probably have a general applicability for a wide variety of image-classification applications.
Conference Paper
The genetic programming (GP) method is proposed as a new approach to perform texture classification based directly on raw pixel data. Two alternative genetic programming representations are used to perform classification. These are dynamic range selection (DRS) and static range selection (SRS). This preliminary study uses four brodatz textures to investigate the applicability of the genetic programming method for binary texture classifications and multi-texture classifications. Results indicate that the genetic programming method, based directly on raw pixel data, is able to accurately classify different textures. The results show that the DRS method is well suited to the task of texture classification. The classifiers generated in our experiments by DRS have good performance over a variety of texture data and offer GP as a promising alternative approach for the difficult problem of texture classification.
Conference Paper
Five alternative methods are proposed to perform multi-class classification tasks using genetic programming. These methods are: (1) binary decomposition, in which the problem is decomposed into a set of binary problems and standard genetic programming methods are applied; (2) static range selection, where the set of real values returned by a genetic program is divided into class boundaries using arbitrarily-chosen division points; (3) dynamic range selection, in which a subset of training examples are used to determine where, over the set of reals, class boundaries lie; (4) class enumeration, which constructs programs similar in syntactic structure to a decision tree; and (5) evidence accumulation, which allows separate branches of the program to add to the certainty of any given class. The results show that the dynamic range selection method is well-suited to the task of multi-class classification and is capable of producing classifiers that are more accurate than the other methods tried when comparable training times are allowed. The accuracy of the generated classifiers was comparable to alternative approaches over several data sets
Conference Paper
This paper presents investigations with digital spectral and textural classification of urban environment using SPOT XS imagery and grey level co-occurrence matrix (GLCM) approach. The results indicate that a combination of texture and spectral features significantly improves the classification accuracy compared to a classification with pure spectral features only. This improvement is about 9% for an addition of one texture feature
Conference Paper
This paper proposes an efficient parallel approach to texture classification for image retrieval. The idea behind this method is to pre-extract texture features in terms of texture energy measurement associated with a `tuned' mask and store them in a multi-scale and multi-orientation texture class database via a two-dimensional linked list for query. Thus each texture class sample in the database can be traced by its texture energy in a two-dimensional row sorted matrix. The parallel searching strategies are introduced for fast identifying the entities closest to the input texture throughout the given texture energy matrix. In contrast to the traditional search methods, our approach incorporates different computation patterns for different cases of available processor numbers and concerns with robust and work-optimal parallel algorithms for row-search and minimum-find based an the accelerated cascading technique and the dynamic processor allocation scheme. Applications of the proposed parallel search and multisearch algorithms to both single image classification and multiple image classification are discussed. The time complexity analysis shows that our proposal will speed up the classification tasks in a simple but dynamic manner. Examples are presented of the texture classification task applied to image retrieval of Brodatz textures, comprising various orientations and scales
Conference Paper
Quality control in magnetic disk production is a crucial issue, in particular the condition of the head disk `interface' which is critical to the overall quality of the final product. The verification of this `interface' has been complicated by advances in technology which have seen the height at which the head glides above the disk, the fly-height, reduce to micron order. This has necessitated inspection of the disks for defects of comparable magnitude. There is a growing need for visual inspection which is low cost and computationally fast in order to achieve total quality control. The micrograph of the disk surface is a natural texture and, whilst defect boundaries can be perceived by human inspectors, they do not correspond to a local discontinuity in intensity. Conventional edge detection operators are therefore unsuitable for segmentation. In this paper, segmentation is achieved via texture recognition. A texture recognition method based on a co-occurrence spectrum of n<sup>th</sup> order is used and defect boundaries are defined by discontinuities in the texture within a window scanning the image
Conference Paper
A radar image of the land or sea will contain extended regions of apparently homogeneous textures arising from, for example, forested areas or uniformly driven expanses of water. However, such regions may contain subtle texture anomalies which indicate the presence of a feature of interest such as a change in tree density due to environmental effects, damping of the sea surface due to the presence of pollutants or the presence of land-based or maritime military targets. For this reason, it is of interest to develop methods of detecting anomalies in textured regions. A strategy for texture anomaly detection is to estimate the multi-variate probability density function (PDF) of a texture segment using the whole homogeneous region as training data and then to calculate the likelihood of occurrence of each individual segment within the region given the PDF estimate. Those texture segments with anomalous likelihood values can then be flagged as potential texture anomalies