Zheru Chi

Handwritten numeral recognition using self-organizing maps and fuzzy rules

Abstract Handwritten numeral recognition using combined self-organizing maps (SOMs) and fuzzy rules is presented in this paper. In the learning phase, the SOM algorithm is used to produce prototypes which together with corresponding variances are used to determine fuzzy regions and membership functions. Fuzzy rules are then generated by learning from training patterns. In the recognition stage, an input pattern is classified by a fuzzy rule based classifier. An unsure pattern is then re-classified by an SOM classifier. Experiments on a database of 20,852 handwritten numerals (10,426 used for training and a further 10,426 for testing) show that this combination technique achieves satisfactory results in terms of classification accuracy and time, and computer memory required.

A neural root finder of polynomials based on root moments

This letter proposes a novel neural root finder based on the root moment method (RMM) to find the arbitrary roots (including complex ones) of arbitrary polynomials. This neural root finder (NRF) was designed based on feedforward neural networks (FNN) and trained with a constrained learning algorithm (CLA). Specifically, we have incorporated the a priori information about the root moments of polynomials into the conventional backpropagation algorithm (BPA), to construct a new CLA. The resulting NRF is shown to be able to rapidly estimate the distributions of roots of polynomials. We study and compare the advantage of the RMM-based NRF over the previous root coefficient methodbased NRF and the traditional Muller and Laguerre methods as well as the mathematica roots function, and the behaviors, the accuracies of the resulting root finders, and their training speeds of two specific structures corresponding to this FNN root finder: the log and the FNN. We also analyze the effects of the three controlling parameters P0 p with the CLA on the two NRFs theoretically and experimentally. Finally, we present computer simulation results to support our claims.

A robust eye detection method using combined binary edge and intensity information

In this paper, a new eye detection method is presented. The method consists of three steps: (1) extraction of binary edge images (BEIs) from the grayscale face image based on multi-resolution wavelet transform, (2) extraction of eye regions and segments from BEIs and (3) eye localization based on light dots and intensity information. In the paper, an improved face region extraction algorithm and a light dots detection algorithm are proposed for better eye detection performance. Also a multi-level eye detection scheme is adopted. Experimental results show that a correct eye detection rate of 98.7% can be achieved on 150 Bern images with variations in views and gaze directions and 96.6% can be achieved on 564 AR images with different facial expressions and lighting conditions.

Attention-driven image interpretation with application to image retrieval

Visual attention, a selective procedure of humans early vision, plays a very important role for humans to understand a scene by intuitively emphasizing some focused regions/objects. Being aware of this, we propose an attention-driven image interpretation method that pops out visual attentive objects from an image iteratively by maximizing a global attention function. In this method, an image can be interpreted as containing several perceptually attended objects as well as a background, where each object has an attention value. The attention values of attentive objectives are then mapped to importance factors so as to facilitate the subsequent image retrieval. An attention-driven matching algorithm is proposed in this paper based on a retrieval strategy emphasizing attended objects. Experiments on 7376 Hemera color images annotated by keywords show that the retrieval results from our attention-driven approach compare favorably with conventional methods, especially when the important objects are seriously concealed by the irrelevant background.

Leaf Image Retrieval with Shape Features

In this paper we present an efficient two-step approach of using a shape characterization function called centroid-contour distance curve and the object eccentricity (or elongation) for leaf image retrieval. Both the centroid-contour distance curve and the eccentricity of a leaf image are scale, rotation, and translation invariant after proper normalizations. In the frist step, the eccentricity is used to rank leaf images, and the top scored images are further ranked using the centroid-contour distance curve together with the eccentricity in the second step. A thinning-based method is used to locate start point(s) for reducing the matching time. Experimental results show that our approach can achieve good performance with a reasonable computational complexity.

A background-thinning-based approach for separating and recognizing connected handwritten digit strings

Most algorithms for segmenting connected handwritten digit strings are based on the analysis of the foreground pixel distributions and the features on the upper/lower contours of the image. In this paper, a new approach is presented to segment connected handwritten two-digit strings based on the thinning of background regions. The algorithm first locates several feature points on the background skeleton of a digit image. Possible segmentation paths are then constructed by matching these feature points. With geometric property measures, all the possible segmentation paths are ranked using fuzzy rules generated from a decision-tree approach. Finally, the top ranked segmentation paths are tested one by one by an optimized nearest neighbor classifier until one of these candidates is accepted based on an acceptance criterion. Experimental results on NIST special database 3 show that our approach can achieve a correct classification rate of 92.5% with only 4.7% of digit strings rejected, which compares favorably with the other techniques tested.

Document Image Recognition based on template matching of component block projections

Document Image Recognition (DIR), a very useful technique in office automation and digital library applications, is to find the most similar template for any input document image in a prestored template document image data set. Existing methods use both local features and global layout information. In this paper, we propose a novel algorithm based on the global matching of Component Block Projections (CBP), which are the concatenated directional projection vectors of the component blocks of a document image. Compared to those existing methods, CBP-based template-matching methods possess two major advantages: (1) The spatial relationship among the component blocks of a document image is better represented, hence a very high matching accuracy can be obtained even for a large template set and seriously distorted input images; and (2) the effective matching distance of each template and the triangle inequality are proposed to significantly reduce the computational cost. Our experimental results confirm these advantages and show that the CBP-based template-matching methods are very suitable for DIR applications.

Two-stage segmentation of unconstrained handwritten Chinese characters

Correct segmentation of handwritten Chinese characters is crucial to their successful recognition. However, due to many difficulties involved, little work has been reported in this area. In this paper, a two-stage approach is presented to segment unconstrained handwritten Chinese characters. A handwritten Chinese character string is first coarsely segmented according to the background skeleton and vertical projection after a proper image preprocessing. With several geometric features, all possible segmentation paths are evaluated by using the fuzzy decision rules learned from examples. As a result, unsuitable segmentation paths are discarded. In the fine segmentation stage that follows, the strokes that may contain segmentation points are first identified. The feature points are then extracted from candidate strokes and taken as segmentation point candidates through each of which a segmentation path may be formed. The geometric features similar to the coarse segmentation stage are used and corresponding fuzzy decision rules are generated to evaluate fine segmentation paths. Experimental results on 1000 Chinese character strings from postal mail show that our approach can achieve a reasonable good overall accuracy in segmenting unconstrained handwritten Chinese characters.

Leaf Vein Extraction Using Independent Component Analysis

The purpose of this work is to develop an interactive tool which helps botanists to extract the vein system with its hierarchical properties with as little user interaction as possible. In this paper, we present a new venation extraction method using independent component analysis (ICA). The popular and efficient FastICA algorithm is applied to patches of leaf images to learn a set of linear basis functions or features for the images and then the basis functions are used as the pattern map for vein extraction. In our experiments, the training sets are randomly generated from different leaf images. Experimental results demonstrate that ICA is a promising technique for extracting leaf veins and edges of objects. ICA, therefore, can play an important role in automatically identifying living plants.

Zeroing polynomials using modified constrained neural network approach

This paper proposes new modified constrained learning neural root finders (NRFs) of polynomial constructed by backpropagation network (BPN). The technique is based on the relationships between the roots and the coefficients of polynomial as well as between the root moments and the coefficients of the polynomial. We investigated different resulting constrained learning algorithms (CLAs) based on the variants of the error cost functions (ECFs) in the constrained BPN and derived a new modified CLA (MCLA), and found that the computational complexities of the CLA and the MCLA based on the root-moment method (RMM) are the order of polynomial, and that the MCLA is simpler than the CLA. Further, we also discussed the effects of the different parameters with the CLA and the MCLA on the NRFs. In particular, considering the coefficients of the polynomials involved in practice to possibly be perturbed by noisy sources, thus, we also evaluated and discussed the effects of noises on the two NRFs. Finally, to demonstrate the advantage of our neural approaches over the nonneural ones, a series of simulating experiments are conducted.