Showing papers by "Ching Y. Suen published in 1991"
TL;DR: A new method of recognizing handwritten Chinese characters using a structural representation called hierarchical attributed graph representation (HAGR) which can tolerate the variations of HAGR which reflect the instabilities or variabilities of handwrittenChinese characters resulting from different writing styles.
140 citations
TL;DR: A new method called transformation-ring-projection (TRP) is proposed to achieve the size-orientation-invariance characteristic and provides the feasibility or VLSI implementation to speed up computation for real-time processing.
Abstract: The size-orientation-invariance characteristic plays an important role in pattern recognition. It has many applications in computer vision, optical character recognition (OCR), office automation, electronic publication, graphics, etc. In this paper, a new method called transformation-ring-projection (TRP) is proposed to achieve this characteristic. In TRP, shape transformation technique is employed to center the pattern image and normalize its size; the ring-projection scheme is used to handle the orientation problem. An experiment was conducted to verify the proposed method in character recognition. The TRP algorithm requires only simple and regular operations, and provides the feasibility or VLSI implementation to speed up computation for real-time processing. A study on VLSI architecture with extensive parallel processing and pipelining capabilities for the proposed TRP algorithm is also presented.
59 citations
TL;DR: This paper outlines some of the issues involved in evaluating expert systems and cites almost 200 significant papers on the topic and aims to help both new and established researchers become acquainted with the literature of an important and growing field.
Abstract: The rapid growth in the use of expert systems has led to widespread interest in techniques for their evaluation, which includes verification and validation. Evaluation techniques in the software engineering community are well established but there are different opinions about the effectiveness of transferring software engineering methodologies to expert systems. This paper outlines some of the issues involved in evaluating expert systems and cites almost 200 significant papers on the topic. We present it with the expectation that it will help both new and established researchers become acquainted with the literature of an important and growing field.
42 citations
08 Jul 1991
TL;DR: A novel combination technique, called an associative switch, is developed for solving the problem of combining multiple classifiers for recognizing totally unconstrained handwritten numerals using a neural net trained by the backpropagation technique with a modified energy criterion.
Abstract: Explores the possibility of using a neural-net approach to the task of combining multiple classifiers. A combination principle is proposed, and a novel combination technique, called an associative switch, is developed for solving the problem. The switch is controlled by a neural net trained by the backpropagation technique with a modified energy criterion. When an unlabeled pattern is the input to each individual classifier, it also goes to the neural net for associatively calling out a code which controls the switch to decide whether the result of each classifier could pass through as a final result. This associative switch is applied to a problem of combining multiple classifiers for recognizing totally unconstrained handwritten numerals. >
36 citations
TL;DR: Experiments with a wide variety of binary patterns show that this new knowledge-based thinning algorithm generates better skeletons than several other well-known algorithms.
30 citations
TL;DR: Algorithms for improving the quality of skeletons obtained from sequential thinning processes take into consideration the shape of the contours in the dynamic removal and placement of pixels in order to obtain skeletons that are more representative of the shapes of the original patterns.
19 citations
13 Oct 1991
TL;DR: The authors studied several computer algorithms by evaluating the recognition rates on distinct parts of handprinted characters, and the total mean score of recognition rates of parts obtained from the proposed algorithms was 30% higher than that from human experiments.
Abstract: The authors studied several computer algorithms by evaluating the recognition rates on distinct parts of handprinted characters. A regional decomposition method is to facilitate pattern analysis and recognition, by which a complicated pattern can be split into simpler parts or subpatterns. A hierarchy model is also proposed to identify the attributes in different hierarchic levels of patterns so that the recognition rates of the different parts can be calculated by a simplified, uniform, computational scheme. The computed results coincide with those of subjective experiments, but they are more precise and complete. The total mean score of recognition rates of parts obtained from the proposed algorithms was 30% higher than that from human experiments. >
5 citations
TL;DR: Two examples are presented to illustrate how the proposed MLIS solves a practical engineering problem of pattern recognition and how the ERT is applied to design a pattern recognizer.
4 citations
01 Jan 1991
TL;DR: The invention relates to an operational amplifier (OA) or comparator circuit with minimized offset voltage and drift, particularly for analog and/or digital systems requiring high speed and high accuracy signal processing, such as integrators, analog-to-digital converters, etc.
Abstract: The invention relates to an operational amplifier (OA) or comparator circuit with minimized offset voltage and drift, particularly for analog and/or digital systems requiring high speed and high accuracy signal processing, such as integrators, analog-to-digital converters, etc. The provided OA and comparators circuits retain a high speed and have a drastically increased accuracy. No correcting amplifier is necessary. The gain set by external components is insignificant. Widely available high speed switching components can be used. The tolerance of the passive components used is not critical. The values of the capacitors are determined by the switching frequency and amount of the input bias current of the employed OA. A FET OA can be used as its very large offset voltage and drift can be easily compensated.
01 Jan 1991
TL;DR: The pattern classification capability of associative memory network is shown through the recognition of a set of multi-font Chinese characters through the selection of associated class vectors for each pattern prototype being stored.
Abstract: This paper shows the pattern classification capability of associative memory network through the recognition of a set of multi-font Chinese characters. Associative memory can become a suitable pattern classifier by an appropriate selection of the inner codes, or specifically, the selection of associated class vectors for each pattern prototype being stored. In our experiment, Hadamard vectors are selected in view of the features of the original character data, and their effectiveness is displayed by the network's recognition behaviour.
01 Mar 1991
TL;DR: An image transformation method that performs mapping and filling at the same time, while respecting the connectivity of the original image, is proposed and, as a result, the transformations become more consistent and accurate.
Abstract: An image transformation method that performs mapping and filling at the same time, while respecting the connectivity of the original image, is proposed. As a result, the transformations become more consistent and accurate. Its VLSI implementation can reduce the time complexity to O(N/sup 2/) using a uniprocessor, where N is the dimension of the image plane. The algorithms can handle all kinds of images including those of long narrow objects that present problems to other algorithms They also reduce the errors introduced by the order in which rotation and scaling are applied. Their application to gray level images is studied. A series of experiments has been conducted to verify the performance of the proposed algorithms. >
08 Jul 1991
TL;DR: The pairing strategy when Hadamard vectors have been specially assigned to be the inner codes in an associative memory classifier designed to recognize a set of Chinese characters is discussed.
Abstract: Summary form only given. The authors discuss the pairing strategy when Hadamard vectors have been specially assigned to be the inner codes in an associative memory classifier designed to recognize a set of Chinese characters. It is shown that with an appropriate pairing scheme an improvement is achieved in the classification behavior of the associative memory network. >
24 Jun 1991
24 Jun 1991