Jun-ichiro Hayashi

Method for estimating and modeling age and gender using facial image processing

A preliminary questionnaire (enquete) examination of how effectively facial images could be used for gender and age estimations was executed by using 300 different faces and 21 examinees. Wrinkles appearing in the face and the shape and size of the facial parts are selected to model the age and gender estimation in this research based on this enquete. An image processing algorithm for wrinkle modeling is proposed. In addition, a method for making relationships between faces and their keywords is proposed by using latent semantic indexing. An efficient interface for displaying the relationships among keywords and facial images is introduced with 2D/3D-feature space defined by the indexing.

Age and gender estimation from facial image processing

We are researching about an age and gender estimation based on wrinkle texture and color of facial images. A preliminary questionnaire (enquete) examination on how effectively the facial images could be used for gender and age estimations was executed by using 300 different faces and 21 examinees. Wrinkles appearing in the face and the shape and size of the facial parts are selected to model the age and gender estimation in this research basing on this enquete. Basing on this preliminary consideration, an image-processing algorithm for wrinkle modeling, was proposed. In addition, a method for making relationships between facial images and their keywords was proposed by using the latent semantic indexing. An efficient interface for displaying the relationships among keywords and facial images has been introduced with 2D/3D-feature space defined by the indexing.

Age and gender estimation based on wrinkle texture and color of facial images

We are researching about an age and gender estimation based on wrinkle texture and color of facial images. Preliminary questionnaire (enquete) examination of how effectively the facial images could be used for gender and age estimations was executed by using 300 different faces and 21 examinees. Wrinkles appeared in the face and the shape and size of the facial parts are selected to model the age and gender estimation in this research based on this enquete. Based on this preliminary consideration, an image processing algorithm for wrinkle modeling was proposed in this paper. In addition, a method for making relationships between facial images and their keywords was proposed by using the latent semantic indexing. An efficient interface for displaying the relationships among keywords and facial images has been introduced with 2D/3D-feature space defined by the indexing.

A development of computer-aided diagnosis system using fundus images

We developed a computer-aided diagnosis (CAD) system. The purpose of this project is to aid physicians in diagnosing medical images. A CAD system can help physicians by displaying useful information such as the location of abnormalities. Although many kinds of images are used in medical diagnosis, we have focused our research to methods of analyzing fundus images. Physicians normally observe the condition of blood vessels and the retina when examining fundus images. Therefore, we are researching analysis schemes for these objects. Presently, we have improved schemes to detect abnormalities on the retina and analyze blood vessel intersections. In addition, we have developed methods of detecting abnormal widths in blood vessels. We have collected about 450 fundus images and used them to test our schemes. We have obtained relatively good results. In this paper, we give an outline of our system and describe each processing step.

Robot system for cloth handling

Flexible objects such as clothes are hard to be handled by robots. But the working conditions for laundry factories are severe because of dusts, heats and steams. To automate these factories, the robot system with 3-D Vision systems and a flexible hand are dispensable. To handling the flexible clothes, it is required to recognize the 3-D shape of flexible clothes. To handle the clothes, a robot hand with high freedom of degree should be introduced. Here, 3-D robot vision systems with stereo cameras and a pattern projector have been developed to apply the cloth handling, and a cloth handling robot system with a flexible hand has been developed. In this paper, the experimental development of the cloth handling robot system has been introduced.

Age and Gender Estimation Based on Facial Image Analysis

This paper is focused on the extraction of age and gender features appeared on the face as a semantic features of the face. It is widely known that the number of wrinkles increases and the shapes of the wrinkles are deepen according to the increase of the age. We extracted wrinkles and modelled by a set of line segments extracted by a special Hough transform DTHT for a method for line segment detection, and analized the wrinkle line segments at the corner of eyes. And we experimented it by using HOIP face database consisted of 300 faces. Further more, since there are several differences in facial expressions between male and female, we took the relationship among face and facial parts into account for gender estimation. In this paper, we constructed a method for age and gender estimations and some experimental results were presented. It was known that texture analysis of the wrinkles could provide 68 % success rate (cf. 86% of human performance) for gender estimation and that the shape features of the facial parts could provide around 20% success rate (cf. 30% of human performance) for age estimation.

Facial caricaturing system PICASSO with emotional motion deformation

Discusses how motion features of emotional expression processes could enforce facial caricaturing. A still caricature is not enough to express individual character. It was experimentally known that there are various kinds of motion features in the laughing processes. We implemented a system to deform the motion features caused by emotional expression for enforcing facial caricaturing. We propose an idea to extend the extraction mechanism of the individuality features to the time-axis by means of the in-betweening method. It is clarified that individual facial features can be enforced by emotional motion deformation in caricaturing.

A method for bicycle detection using ellipse approximation

We proposes a method for bicycle detection using ellipse approximation. In this research, we focus on the area of bicycle tire. Bicycle tire is the smallest part of the change in shape of bicycle parts. We detected a bicycle by using ellipse approximation to estimate the tire of the bicycle. In this paper, we describe a method for estimating result of region detection and bicycle tires.

Design of cloth handling robot system

Flexible objects such as clothes are hard to be handled by robots. But the working conditions for laundry factories are severe because of dusts, heats and steams, and full automation systems which can handle clothes are highly required. To automate these factories, the combination of 3-D Vision systems and flexible handling mechanism are dispensable to handle flexible clothes. But they were not so successful. Here, the cloth handling task has been analyzed and the best combination of the sensors and the mechanisms has been pursued. The experimental system has been developed and proved its efficiency.

Evaluation of sensitivity strength for visual inspection system of color display device

To develop the visual inspection systems, it requires long and severe effort of image processing experts. But in many cases, human operators can detect the defects, very easily. “Mura” is a color uniformity defect appeared on display devices such as LCDs or CPTs. It is a vague color spot which appears on the display devices. The detection of the defects depends on human sensitivity for color, size and edge sharpness of the defects. It is obvious that the human sensitivity factors should be introduced to evaluate the visual defects of such devices. In this research, the word “sensitivity strength” has been introduced. The human sensitivity of visual inspection is described by several hypothesizes, and the “sensitivity strength” is evaluated from many experiences. After the evaluation, using these hypothesizes, a human sensitivity simulator has been developed and evaluated its inspection ability.