A new method for restoring high-resolution binary images is presented to improve legibility of low-resolution document images. The initially restored image is generated by simple techniques, and is then improved by integrating a variety of features obtained through image analysis. Missing strokes of characters are complemented based on topographic features. Contours of characters are then modified in terms of gradient magnitudes and curvatures along the contours. Finally, contours are beautified so that they look good to the human eye. The proposed method can deal with characters having complex structures such as Kanji, and entails relatively simple computation. Smoothness and linearity along contours are significantly improved and strokes are restored correctly. The proposed method also improves OCR accuracy.

This paper describes a new approach to restoring scanned color document images where the backside image shows through the paper sheet. A new framework is presented for correcting show-through components using digital image processing techniques. First, the foreground components on the front side are separated from the background and backside components through locally adaptive binarization for each color component and edge magnitude thresholding. Background colors are estimated locally through color thresholding to generate a restored image, and then corrected adaptively through multiscale analysis along with comparison of edge distributions between the original and the restored image. The proposed method does not require specific input devices or the backside to be input; it is able to correct unneeded image components through analysis of the front side image alone. Experimental results are given to verify effectiveness of the proposed method.

This paper describes a new approach to adaptive digital halftoning with the least squares model-based (LSMB) method. A framework is presented for the adaptive control of smoothness and sharpness of the halftone patterns according to local image characteristics. The proposed method employs explicit, quantitative models of the human visual system represented as 2-D linear filters (eye filters). In contrast with the standard LSMB method where the single eye filter is employed uniformly over the image, the model parameters are controlled according to local image characteristics for each pixel. Because of the adaptive selection of eye filters for the pixels, image enhancement is incorporated into the halftoning process. Effectiveness of the proposed approach is demonstrated through experiments using real data compared with the error-diffusion algorithm and the standard LSMB method.

This paper describes a new approach to adaptive digital halftoning with the least squares model-based (LSMB) method. A framework is presented for the adaptive control of smoothness and sharpness of the halftone patterns according to local image characteristics. The proposed method employs explicit, quantitative models of the human visual system represented as 2D linear filters (eye filters). In contrast with the standard LSMB method where the single eye filter is employed uniformly over the image, the model parameters are controlled according to local image characteristics for each pixel. Because of the adaptive selection of eye filters for the pixels, image enhancement is incorporated into the halftoning process. Effectiveness of the proposed approach is demonstrated through experiments using real data compared with the error-diffusion algorithm and the standard LSMB method.

In this paper, a method for recognizing plant species by the shapes of the leaves is presented. The recognition is based on a hierarchical representation of shape features. First, the structures of the shapes of leaves are analyzed and the shapes of components are extracted in detail. The structures and detailed shapes of leaves are approximated by polygons whose vertices are critical points of curvature of contours of leaves. The method also takes variations of shapes of leaves of the same species into account. Experimental results using a number of species in the family of Acer (Maple) indicate that structures of leaves and shapes of their components can be considered as landmarks to recognize the species of plants.

A novel method is presented for extracting closed boundaries of document components such as characters and symbols directly from gray-scale document images based on the surface data structures along with structural features. The method is based on the simple model assuming that a closed boundary of document components can be approximated as a series of horizontal line segments and can be extracted by linking surface components with steep gradients which share commonly intersecting horizontal planes. The proposed algorithm is compared with some binarization algorithms, shown to be effective for improving recognition accuracy for very poor quality data.

The combination of analysis and synthesis is a powerful method in tackling complex problems such as handwriting recognition. Some methods have been proposed for synthesis of deformed patterns for handwritten characters. However, existing methods have some limitations that: 1) specification of a number of standard, initial shapes is required for synthesis of various patterns, or 2) shapes with complex curve structures cannot be handled. On the other hand, for coping with complex deformations of handwritten characters, the integration of structural descriptions and geometrical statistical informations has been shown to be powerful in shape description and matching. Applying this idea to pattern synthesis, we present an algorithm for synthesis of deformed patterns by combining structural descriptions of shapes and curve generation techniques. Based on the shape model (class descriptions) in terms of the high level features integrating several types of informations, we can synthesize various patterns incorporating structural, statistical, geometrical deformations without resorting to some primitive techniques in image transformation or specifying standard, initial patterns.

Presents an approach to automatic construction of structural models incorporating discontinuous transformations, with emphasis on application to unconstrained handwritten character recognition. The author considers this problem as constructing inductively, from the data set, some shape descriptions that tolerate certain types of shape transformations. The approach is based on the exploration of complete, systematic, high-level models on the effects of the transformations, and the generalization process is controlled and supported by the high-level transformation models. An analysis of the a priori effects of commonly occurring discontinuous transformations is carried out completely and systematically, leading to a small, tractable number of distinct cases. Based on this analysis, an algorithm for the inference of super-classes under these transformations is designed. Furthermore, through examples and experiments, the author shows that the proposed algorithm can generalize unconstrained handwritten characters into a small number of classes, and that one class can represent various deformed patterns.

Computer-aided diagnosis of occlusal disorders and design of dental restorations requires an automated evaluation of jaw occlusion and chewing ability. This requires simulation of the motion of the jaws and characterization of contacts between the surfaces of teeth. We propose approaches to evaluation of the load on teeth and of the grinding process. These characteristics are derived in interactive time, and are based on distance maps and topological structure of the contact zones. The proposed approaches are general and usable in applications where modeling of contact between objects with complex geometry is required.>

The clues for discontinuous structural transformation in handwriting lie in deformation around singular points and stroke connections of handwriting. For designing an algorithm for automatic construction of structural models incorporating discontinuous transformations, we analyze systematically how global structures and features are changed by stroke connections. Based on the set of systematic laws of shape transformations, we design an algorithm for finding component correspondence between a pair of deformed patterns, and present an algorithm for automatic construction of structural models for unconstrained handwritten characters along with some examples. The proposed algorithm can generalize unconstrained handwritten characters into a small number of classes, and one class can represent various deformed patterns.

An essential problem in online handwriting recognition is the shape variation along with the variety of stroke number and stroke order. In this paper we present a clear and systematic approach to shape matching based on structural feature grouping. To cope with topological deformations caused by stroke connection and breaking, we incorporate some aspects of top-down approaches systematically into the shape matching algorithm. The grouping of local structural features into high-level features is controlled by high-level knowledge as well as the simple geometric conditions. The shape matching algorithm has the following properties from the viewpoint of online character recognition: (1) stroke order, direction, and number are free, and (2) stroke connection and breaking are allowed.>

An essential problem in online handwriting recognition is in the shape variation along with the variety of stroke number and stroke order. This paper presents a clear and systematic approach to shape matching based on structural feature grouping. To cope with topological deformations caused by stroke connection and breaking, we incorporate some aspects of top-down approaches systematically into the shape matching algorithm. The grouping of local structural features into high-level features is controlled by high-level knowledge as well as the simple geometric conditions. The shape matching algorithm requires a small number of prototypes and has the following properties from the viewpoint of online character recognition: (a) The order of strokes is free; (b) The number of strokes is free; (c) Stroke connection and breaking are allowed.

Recognition of handwritten character strings is a challenging problem, because we need to cope with variations of shapes and touching/breaking of characters at the same time. A natural approach to recognizing such complex objects is as follows: The object is decomposed into segments, and meaningful partial shapes (shapes which are recognized as some characters) are constructed by merging segments locally. Then, a globally consistent interpretation of the object is determined from the combination of partial shapes. This approach can be referred to as a model-based split-and-merge method. Based on this idea, we present an algorithm for recognition and segmentation of character strings. We give systematic performance statistics by experiments using handwritten numerals. This algorithm can be applied to character strings composed of any number of characters and any type of touching or breaking, whether the number of constituent characters is known or unknown.

H. Nishida and S. Mori (1992) proposed a clear, rigorous, and powerful method for the structural description of character shapes in terms of quasi-topological features (convexity and concavity), directional features, and singular points (branch points and crossings). Shapes are described by few components with rich features and shape prototypes (structural models) can be constructed automatically from the training data. However, the number of directions is fixed to 4, and more directions such as 8 or 16 cannot be dealt with. For various applications of Nishida and Mori's method, the author presents a method for structural analysis and description of simple arcs or closed curves based on 2/sup m/ (m = 2,3,4,...) directional features and convex/concave features. Software OCR systems without specialized hardware have attracted much attention recently. Based on the method of structural analysis and description, the author describes a software implementation of a handwritten character recognition system with a multi-stage strategy.>

We present algebraic approach to the inductive learning of structural models and automatic construction of shape prototypes for character recognition on the basis of the algebraic description of curve structure proposed by Nishida and Mori (1991, 1992). A class in the structural models is a set of shapes that can be transformed continuously to each other. We consider an algebraic representation of continuous transformation of components of the shape, and give specific properties satisfied by each component in the class. The generalization rules in the inductive learning are specified from the viewpoints of continuous transformation of components and relational structure among the components. The learning procedure generalizes a pair of classes into one class incrementally and hierarchically in terms of the generalization rules. We show experimental results on handwritten numerals.>

The authors propose a compact and concise method of describing curves in terms of the quasi-topological features and the structure of each singular point. The quasi-topological features are the convexity, loop, and connectivity. The quasi-topological structure is analyzed in a hierarchical way, and algebraic structure is presented explicitly for each representation level. The lower-level representations are integrated into the higher-level one in a systematic way. When a curve has singular points (branch points), the curve is decomposed into components, where each is a simple arc or a simple closed curve, by decomposing each singular point. The description scheme is applied to character recognition.>