Ryoji Haruki

Two-dimensional extension of nonlinear normalization method using line density for character recognition

Some of the authors previously proposed a few nonlinear normalization methods using line density for handwritten characters and it was verified that those nonlinear methods were more effective than a linear method. Those nonlinear methods, however, could not use two-dimensional information of the line density, because the line density was projected onto the axes of coordinates. The paper proposes an extended nonlinear normalization method which uses two-dimensional information of the line density. It is confirmed that the proposed method is more powerful than conventional methods by experiments tested on the handwritten character database ETL-8.

Automatic seal verification using three-dimensional reference seals

The most important problem on automatic seal impression verification is to absorb the various impression quality depending on the various affixing conditions. Over the past years, many studies have been made on absorbing the various impression quality using a two-dimensional image as reference and input. But it is difficult to verify each of a large number of input patterns by a sole two-dimensional reference pattern. This paper proposes a seal identification method using a three-dimensional image (range image) as reference. As results of the verification with actual seal impressions, the verification rate is improved from that by using two-dimensional reference impressions.

Three-dimensional approach to thresholding seal impressions

Most of the conventional methods for automatic seal identification performed one-to-one matching between the reference impression and the input one which are measured as 2D images. However, those methods cannot absorb the various quality by only the matching techniques. This paper proposes a new approach to deal with the various quality of impressions by measuring the seals for reference as 3D (three-dimensional) images, that is, range images. By storing the seal impressions as 3D images, it becomes possible to construct various 2D impressions from only one 3D reference image; according to the affixing conditions of an input impression such as the slant, the pressure, and the state of ink on the seal. Primarily, this paper considers how to construct 2D images from a 3D reference image by determining the affixing conditions. The performance is verified by experiment, and the result shows that plural 2D impressions according to the affixing conditions can be constructed from only one 3D reference image.

A multi-stage algorithm of extracting joint points for generating function-fonts

An automatic function-font generating system for brush-written characters has been proposed. The system extracts sharp corners of boundary lines, (so-called joint points), and approximates the boundary lines between adjacent joint points using a piecewise polynomial. The quality of the generated function-fonts is controlled by the location of joint points. An attempt is made to develop an advanced function-font generating system with an algorithm to extract appropriate joint points. The algorithm consists of three stages: (1) extracting joint points by evaluating digital curvature, (2) extracting additional joint points by evaluating the curvature of approximated boundary line, and (3) removing the redundant joint points. The performance of the method is verified by applying it to brush-written characters, printed characters, and figures.<<ETX>>

Scalable image coding by spline approximation for a gray-scale image

The proposed method expresses a gray-scale image by parametric spline functions for edge components and by two-variable spline functions for low frequency components. It can reconstruct the image keeping its quality for the basic shape transformation. If a binary image is input as a special case, the proposed method can make a scalable vector font automatically. The performance of the proposed method is verified by some experiments.

Observation method for mathematical graphology

An observation method for mathematical graphology is proposed. The motivation of mathematical graphology is to realize graphological information processing on a machine. Furthermore, mathematical graphology aims to construct a new information theory for visual communication. In order to construct the mathematical model this paper focuses on the observation process, which is the first stage in the model. Based on the physiological saccade phenomena, feature points are extracted which include the important information. Then, contours between adjacent feature points are approximated using smooth functions. The performance of the proposed method is verified by experiments.

Data fitting by spline functions using the biorthonormal basis of the B-spline basis

The basis for the spline function space which is orthogonal to the B-spline basis is derived, i.e. the biorthonormal basis of the B-spline basis. The derived basis makes it easy to obtain B-spline coefficients from a given function in a straightforward way. This is applied to data fitting by spline functions, which uses the least square approximation. Then computing quantities and errors are examined, and the effectiveness and potential of this approach is described.

Improved and extended scalable image coding by spline approximation for a binary, gray-scale and color image

The image coding method for a gray-scale image, which can reconstruct the high quality image against several transformations such as scaling by expressing the image signal as the shape of the mathematical function, are proposed. This method is also applicable to a binary image, but it has not been applicable to a color image; there are some problems of quality with this method. The paper attempts to improve the quality and to extend the preview method to a color image. For the quality improvements, the processes involved in region segmentation and approximation for low frequency components are reformed. For the extension to a color image, the process of region segmentation is particularly reformed again. The performance of the proposed method is verified by some experiments. Although the quality problems remain to a certain extent, it is much better than previous results. The extension to a color image proves successful. Moreover, this improvement and extension can parallelize the process more effectively.

On a method of noise reduction for 3D stamped character recognition using fluency-wavelet

A noise reduction method for recognizing stamped characters is proposed. A range image was observed using a range finder system. The recognition rate was estimated using the MAP (maximum a posteriori) matching method. A maximum recognition rate of 85.2% was obtained. The observation using a range finder system has the advantages that stamped characters, which have three-dimensional information, are observed as three-dimensional information, and that the character size is measured as the real size. The specific noise of the range image can be reduced using a fluency-wavelet. In the present work, it is tested for a fluency function of degree 2 as the integral kernel.<<ETX>>