Yu Tao’s research while affiliated with Hong Kong Baptist University and other places

In this paper, a novel approach to feature extraction with wavelet and fractal theories is presented as a powerful technique in pattern recognition. The motivation behind using fractal transformation is to develop a high-speed feature extraction technique. A multiresolution family of the wavelets is also used to compute information conserving micro-features. In this study, a new fractal feature is reported. We employed a central projection method to reduce the dimensionality of the original input pattern, and a wavelet transform technique to convert the derived pattern into a set of subpatterns, from which the fractal dimensions can readily be computed. The new feature is a measurement of the fractal dimension, which is an important characteristic that contains information about the geometrical structure. This new scheme includes utilizing the central projection transformation to describe the shape, the wavelet transformation to aid the boundary identification, and the fractal features to enhance image discrimination. The proposed method reduces the dimensionality of a 2-D pattern by way of a central projection approach, and thereafter, performs Daubechies' wavelet transform on the derived 1-D pattern to generate a set of wavelet transform subpatterns, namely, curves that are non-self-intersecting. Further from the resulting non-self-intersecting curves, the divider dimensions are computed with a modified box-counting approach. These divider dimensions constitute a new feature vector for the original 2-D pattern, defined over the curve's fractal dimensions. We have conducted several experiments in which a set of printed Chinese characters, English letters of varying fonts and other images were classified. Based on the formulation of our new feature vector, the experiments have satisfying results.

In this paper, a novel approach to feature extraction based on fractal theory is presented as a powerful technique in pattern recognition. This paper presents a new fractal feature that can be applied to extract the feature of two-dimensional objects. It is constructed by a hybrid feature extraction combining wavelet analysis, central projection transformation and fractal theory. New fractal feature and fractal signatures are reported. A multiresolution family of the wavelets is also used to compute information conserving micro-features. We employed a central projection method to reduce the dimensionality of the original input pattern. A wavelet transformation technique to transform the derived pattern into a set of sub-patterns. Its fractal dimension can readily be computed, and to use the fractal dimension as the feature vectors. Moreover, a modified fractal signature is also used to distinguish the distinct handwritten signatures. We expect that the proposed fractal method can also be used for improving the extraction and classification of features in pattern recognition.

In this paper, a new method of feature extraction with rotation invariant property is presented. One of the main contributions of this study is that a rotation invariant signature of 2-D contours is selected based on the fractal theory is proposed. The rotation invariant signature is a measure of the fractal dimensions, which is rotation invariant based on a series of central projection transform (CPT) groups. As the CPT is applied to a 2-D object, a unique contour is obtained. In the unfolding processing, this contour is further spread into a central projection unfolded curve, which can be viewed as a periodic function due to the different orientations of the pattern. We consider the unfolded curves to be non-empty and bounded sets in IR n , and the central projection unfolded curves with respect to the box computing dimension are rotation invariant. Some experiments with positive results are conducted. This approach is applicable to a wide range of areas such as image analysis, pattern recognition, etc. 1.

In this paper, we are investigating the utility of several emerging techniques to extract features. A novel method of feature extraction is proposed, which includes utilizing the central projection transformation (CPT) to describe the shape, the wavelet transformation to aid in the boundary identification, and the fractal features to enhance image discrimination. It reduces the dimensionality of a two-dimensional pattern by way of a central projection approach, and thereafter, performs Daubechies' wavelet transform on the derived one-dimensional pattern to generate a set of wavelet transform sub-patterns, namely, curves that are non-self-intersecting. The divider dimensions are computed from these curves with a modified box-counting approach. These divider dimensions constitute a new feature vector for the original two-dimensional pattern, defined over the curve's fractal dimensions. We have conducted several experiments in which a set of printed Chinese characters, English letters of varying fonts and other images were classified. Based on the Euclidean distance between the different feature vectors, the experiments have satisfying results.

An approach to feature extraction based on fractal theory is presented as a powerful technique in pattern recognition. It can be used to extract the features of 2D objects, and identify different scripts. A fractal feature and fractal signature are reported. A multiresolution family of the wavelets is also used to compute information conserving micro-features. We employed a central projection method to reduce the dimensionality of the original input pattern, and a wavelet transformation technique to transform the derived pattern into a set of sub-patterns, from which the fractal dimension can readily be computed. Moreover, we have proposed an approach to classify different language using the modified fractal signature. For all these cases, difference in fractal dimension can yield the significative values. We expect that the proposed fractal method can also be used for improving the extraction and classification of features in a pattern recognition system

A new method called central projection transformation is proposed in this paper for feature extraction. From our experiments, the new method is found to be efficient in extracting features from Chinese characters, which contain a vast amount of information. Chinese characters have complex structures, and some of them are composed of several separate components, so several contours are embedded in a character. This may obstruct application of the contour approach to recognizing Chinese characters. Central projection transformation can convert such a multicontour pattern into a solid convex pattern, whose contour is a unique polygon. Most of the information of this new pattern is still located around its peripheries. In this paper, information contents and entropy measurements are studied in both original Chinese characters and transformed new objects from the 3500 most frequently used Chinese characters. The results indicate that both the information contents and entropy measurements of pixels vary according to the positions of the points, and that most of the information is located around the peripheries of the original characters as well as of the new ones. This approach can greatly simplify the processing of Chinese characters and other multicontour patterns. It is also a powerful tool for processing Arabic characters, Japanese characters and other characters.

Proposes a method that reduces the dimensionality of a 2D pattern by means of a central projection approach, and thereafter performs a Daubechies wavelet transformation on the derived 1D pattern to generate a set of wavelet transformation sub-patterns, namely curves that are non-self-intersecting. Further, from the resulting non-self-intersecting curves, the divider dimensions are compared with the modified box-counting approach. These divider dimensions constitute a new feature vector for the original 2D pattern, defined over the curve's fractal dimensions

As the interest in fractal geometry rises, the applications are getting more and more numerous in many domains. This paper deals with the problem of recognizing and classification to optical character recognition. For this purpose, we present a new method of feature extraction based on the principles of fractal geometry and wavelet. This allows us to establish a classification of Chinese character in order to apply to each of the isolated categories the most adapt recognition methods. In particular, the proposed method reduces the dimensionality of a two-dimensional pattern by way of a central projection approach, and thereafter, performs Daubechies' wavelet transformation on the derived one-dimensional pattern to generate a set of wavelet transformation sub-patterns, namely, curves that are non-self-intersecting. Further from the resulting non-self-intersecting curves, the divider dimensions are computed with modified box-counting approach. These divider dimensions constitute a new feature vector for the original two-dimensional pattern, defined over the curve's fractal dimensions. We have conducted several experiments in which a set of printed alphanumeric symbols and Chinese characters of varying fonts and orientation were classified, based on the formulation of our new feature vector. The results obtained from these experiments have consistently shown the character recognition method with the proposed feature vector can yield an excellent classification rate 100%.© (1999) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

As the interest in fractal geometry rises, the applications are getting more and more numerous in many domains. This paper deals with the problem of recognizing and classification to optical character recognition. For this purpose, we present a new method of feature extraction based on the principles of fractal geometry and wavelet. This allows us to establish a classification of Chinese character in order to apply to each of the isolated categories the most adapt recognition methods. In particular, the proposed method reduces the dimensionality of a two-dimensional pattern by way of a central projection approach, and thereafter, performs Daubechies' wavelet transformation on the derived one-dimensional pattern to generate a set of wavelet transformation sub-patterns, namely, curves that are non-self-intersecting. Further from the resulting non-self-intersecting curves, the divider dimensions are computed with modified box-counting approach. These divider dimensions constitute a new feature vector for the original two-dimensional pattern, defined over the curve's fractal dimensions. We have conducted several experiments in which a set of printed alphanumeric symbols and Chinese characters of varying fonts and orientation were classified, based on the formulation of our new feature vector. The results obtained from these experiments have consistently shown the character recognition method with the proposed feature vector can yield an excellent classification rate 100%.