Yoshihiro Mitani

A gabor filter-based method for recognizing handwritten numerals

Abstract We study a Gabor-filter-based method for handwritten numeral character recognition. The Gabor filter is based on a multi-channel filtering theory for processing visual information in the early stages of the human visual systems. The performance of the Gabor-filter-based method is demonstrated on the ETL-1 database. Experimental results show that the artificial neural-network classifier achieved the error rate of 2.34% for a test set of 7000 characters. Therefore, the Gabor-filter-based method should be considered in recognition of handwritten numeric characters.

A local mean-based nonparametric classifier

A considerable amount of effort has been devoted to design a classifier in practical situations. In this paper, a simple nonparametric classifier based on the local mean vectors is proposed. The proposed classifier is compared with the 1-NN, k-NN, Euclidean distance (ED), Parzen, and artificial neural network (ANN) classifiers in terms of the error rate on the unknown patterns, particularly in small training sample size situations. Experimental results show that the proposed classifier is promising even in practical situations.

A Method for Crack Detection on a Concrete Structure

Recently, interest in automatic crack detection on concrete structure images for non-destructive inspection has been increasing. In general, there are various noises such as irregularly illuminated conditions, shading, blemishes and divots in the concrete images. These lead to difficulties for automatic crack detection. This paper presents two p re-processings in order to remove such noises for crack detection. First, slight variations like irregularly illuminated conditions and shading are removed from concrete images by the subtraction pre-processing with the smoothed image. Secondly, a line filter based on the Hessian matrix is used to emphasize line structures associated with cracks. Finally, thresholding processing is used to separate cracks from background. The performance of the proposed method is evaluated by ROC analysis with 50 real images. The experimental results show that the proposed method is effective for detecting cracks on noisy concrete images

A Method for Reading a Resistor by Image Processing Techniques

The resistance of a resistor is defined by colored lines printed on the resistors body. Normally, people read it by sight. Though, if a computer performs this instead, we can reduce the costs. In this paper, we propose a method for reading this resistance by image processing techniques. We extract colors from a real resistors picture, and classify it by its colors. The experimental results show the effectiveness of the proposed method.

Evaluation of an anti-regularization technique in neural networks

An anti-regularization technique which has been recently proposed by Raudys (1995) is studied in small training sample size situations. Experimental results show that as long as the weights of a network are initialized in a very narrow interval, the anti-regularization technique offers significant advantages in terms of both the generalization ability and learning time.

Use of bootstrap samples in designing artificial neural network classifiers

We propose a new bootstrap method for designing artificial neural network (ANN) classifiers. Moreover, the classification performance of ANN classifiers based on the new bootstrap method is demonstrated in small training sample size situations on the artificial data sets.

Training sample selection based on self-training for liver cirrhosis classification using ultrasound images

Ultrasound imaging is a popular and non-invasive tool used in the diagnoses of liver disease. Cirrhosis is a chronic liver disease and it can advance to liver cancer. Early detection and appropriate treatment are crucial to prevent liver cancer. However, ultrasound image analysis is very challenging, because of the low signal-to-noise ratio of ultrasound images. To achieve the higher classification performance, selection of training regions of interest (ROIs) is very important that effect to classification accuracy. The purpose of our study is cirrhosis detection with high accuracy using liver ultrasound images. In our previous works, training ROI selection by MILBoost and multiple-ROI classification based on the product rule had been proposed, to achieve high classification performance. In this article, we propose self-training method to select training ROIs effectively. Evaluation experiments were performed to evaluate effect of self-training, using manually selected ROIs and also automatically selected ROIs. Experimental results show that self-training for manually selected ROIs achieved higher classification performance than other approaches, including our conventional methods. The manually ROI definition and sample selection are important to improve classification accuracy in cirrhosis detection using ultrasound images.

A liver cirrhosis classification on B-mode ultrasound images by the use of higher order local autocorrelation features

In this paper, in order to classify liver cirrhosis on regions of interest (ROIs) images from B-mode ultrasound images, we have proposed to use the higher order local autocorrelation (HLAC) features. In a previous study, we tried to classify liver cirrhosis by using a Gabor filter based approach. However, the classification performance of the Gabor feature was poor from our preliminary experimental results. In order accurately to classify liver cirrhosis, we examined to use the HLAC features for liver cirrhosis classification. The experimental results show the effectiveness of HLAC features compared with the Gabor feature. Furthermore, by using a binary image made by an adaptive thresholding method, the classification performance of HLAC features has improved.

Training ROI Selection Based on MILBoost for Liver Cirrhosis Classification Using Ultrasound Images

Ultrasound images are widely used for diagnosis of liver cirrhosis. In most of liver ultrasound images analysis, regions of interest (ROIs) are selected carefully, to use for feature extraction and classification. It is difficult to select ROIs exactly for training classifiers, because of the low SN ratio of ultrasound images. In these analyses, training sample selection is important issue to improve classification performance. In this article, we have proposed training ROI selection using MILBoost for liver cirrhosis classification. In our experiments, the proposed method was evaluated using manually selected ROIs. Experimental results show that the proposed method improve classification performance, compared to previous method, when qualities of class label for training sample are lower.

Classification of Liver Cirrhosis on m-Mode Ultrasound Images by Extended Higher Order Local Autocorrelation Features

—Ultrasound images are widely used for diagnosis of liver cirrhosis. In liver cirrhosis classification on M-mode ultrasound images, the use of higher order local auto-correlation (HLAC) features has been shown to be effective. In the previous study, we used the traditional 25 dimensional HLAC features. The 25 HLAC features are made by 25 mask patterns with up to 0th, 1st, and 2nd-order. On the other hand, there exists an extension of HLAC features. The extended HLAC features were shown to be more effective when higher-order HLAC features were used. Therefore, by the use of the extended HLAC features, we expected the liver cirrhosis classification performance to improve. However, the effectiveness of the extended HLAC features to classify the liver cirrhosis images is not clear. In this paper, more effectively to classify liver cirrhosis M-mode ultrasound images, we examine the performance of extended HLAC features.