Kazunori Higuchi

Merging multiple views using a spherical representation

This paper proposes a new method for building a 3-D model from a set of range images. The method can merge data of free-form surfaces obtained from arbitrary viewing directions, with no prior knowledge of the poses. Our approach is based on matching the spherical representations of an object between the views. To obtain the spherical representation, we deform a discrete mesh to fit the object surface. A variation of the Gaussian curvature metric, which we call simplex angle, is computed at each node on the deformed mesh and mapped to a coordinate on the unit sphere. The transformation of the objects is computed by comparing the simplex angle measure at each node on the unit sphere. The transformation which produces the minimum errors is selected as the best match. We have implemented this method, applied the method to range images of objects from arbitrary viewpoints, and demonstrated the applicability for modeling from observation.<<ETX>>

Automatic calibration of an in-vehicle gaze tracking system using driver's typical gaze behavior

Many research works have been carried out to measure and use a drivers gaze directions to prevent traffic accidents caused by inattentive driving, neglect to confirm safe conditions, and other driver errors. A calibration process is needed to measure correct gaze directions for a gaze tracking system. However, existing calibration methods require a driver to gaze at specified points before driving. In this paper, we propose a method for automatic calibration of an in-vehicle gaze tracking system by analyzing the drivers typical gaze behavior. The proposed method uses the rear-view and the side-view mirror positions as reference points. The effectiveness of the proposed method is demonstrated by experiments on measuring gaze directions in actual road environments.

Simulating the human vision of elderly for designing control panels

We propose an approach to designing control panels for the elderly based on a computational model of human visual properties, adaptation to light, spatial sensitivity and spectral sensitivity of eyes. The model is implemented as a filter to simulate a human visual system using image processing techniques. The paper describes the proposed method and shows evaluation results. The system allows the designer to predict the difficulties for the elderly to read the visual information under various illumination environments, and we apply our visual model to the prototype evaluation of control panels of the consumer product, including shower-toilets with illuminated buttons.

Displaying method for on-board displays adaptive to luminance of the driving environment

Recently, number of on-board visual display, such as a navigation system is increasing. A driver gazes such display in short period of time in driving. Thus, it is important to improve visibility of the on-board display. We propose two methods about visibility improvement of on-board display. First, we mention about simulating method of creation of the perceived brightness distribution of a scene which is useful for design of the high visibility on-board display. Second, we describe displaying method for the on-board display to control luminance and chromatic saturation of displaying image. Both methods are based on the adaptation property of human vision. Experimental results of subjective and objective evaluation that compare the conventional display and our new one indicate the effectiveness of the second method.

Three-Dimensional Image Information Media. A 3-D Shape Measurement Method for Objects with Uneven Reflective Surfaces.

A method for measuring 3-D objects with surfaces composed of various reflective regions is developed in which the 3-D coordinates of the object sections are obtained by triangulation with the light-section method by using a slit-light source and a TV camera. A slit-light source in which the lighting intensity is controllable and a TV camera with an expanded dynamic range are employed to reduce the influence of uneven reflective surfaces. The TV camera expands its dynamic range by combining multiple images captured at different shutter speeds. A trial 3-D shape measurement system is produced using this method, and its effectiveness is confirmed by experiments.

Recognition of position and orientation of jumbled parts using cross slit light 3-D vision sensor

A vision system and a recognition algorithm to detect position and orientation of jumbled simple-shape parts have been developed. The vision system is composed of a TV camera to detect the two-dimensional position of the parts in a shallow pallet and a recently developed cross slit light three-dimensional vision sensor to detect three-dimensional position and orientation of the parts. In a gripping experiment involving 100 parts, in five shallow pallets, using the vision system and the algorithm, the gripping of jumbled parts with maximum inclination of 30 degrees was successful.<<ETX>>