[in Japanese]

View full abstract

[in Japanese]

View full abstract

A circular cylinder defined by binocular disparity appears flattened at a far distance. It is assumed that the subject estimates the 3D shape using a distorted perception of the physical viewing distance. If the inaccurate distance causes the shape distortion, the estimated lengths for both height and depth would be shorter than the actual lengths. Here, we investigated whether the inaccurate distance could completely account for the flattened perception, estimating the depth and height of the cylinder separately. The results proved that the inaccurate distance does not account for the flattened perception. Instead, the viewing distance affects only the depth perception. Moreover, the following experiment indicated that adding texture and shading cues to the binocular cues does not change the characteristics of the binocular cues only.

View full abstract

The purpose of this study was to examine the fusion area in a real image and binocular disparity when the viewing distance was changed. As a result, when the viewing distance was short, the fusion area in the real image increased. However, the fusion area in the binocular disparity image did not increase as it did in the real image. Since the depth of field with the real image was shallow when the viewing distance was short, it was easy to perceive blurring. In binocular disparity images, since images of the front and back are not blurred, it is impossible to expand the fusion area of binocular disparity, to confirm whether it is expanded by blurring, we measured the fusion area when the binocular disparity image is blurred. As a result, the fusion area of binocular disparity was wider than in the real image when binocular disparity image was blurred.

View full abstract

The proficiency characteristics of a wireless game controller are compared with those of a “MOUSE” in controlling the screens of visual teaching materials that are stored on a PC. The following results were obtained : (1) The wireless game controller could control the screens of PCs within a range of more than 15m in a classroom or in a gymnasium. (2) The proficiency level obtained with the mouse was better than that with a game controller. However, this difference was smaller after some trials. (3) Measurements by nine university students revealed that the average time to search for four visual materials in a three-layered tree ofstructured teaching materials with the game controller was half the average search time required for the mouse when the subject had to walk 5m to his/her PC to control the digital images.

View full abstract

We measured ion-induced secondary electron yield, XRD, and EPMA analysis for oxide compounds including Mg to evaluate the suitability of oxide compounds as protective materials for ac PDPs. The vacuum evaporated thin films of oxide compounds do not retain their original composition on the substrate.

View full abstract

The influence that picture quality differences give to the perceived overall quality of the picture is examined when a stereoscopic picture is of a disparate quality in object and background area, . A subjective assessment test used pictures for evaluation made by encoding several parameters. These pictures showed differences in stereoscopic objects and other background areas. The picture quality of the objects and the background areas was defined based on a the subjective assessment score, and an estimation method for assessing the overall picture quality was examined using the assessment scores. An objective picture quality evaluation model, constructed without using the original picture was also examined.

View full abstract

We propose two methods for improving the character extraction accuracy and for ranking the voice outputs in a system that assists vision impaired people acquire character information using a notebook PC and a CCD camera. We identified the region on a sign board using the color of the area from which characters are extracted by local features and found more accurate character strings from projections on the board region. This reduces the number of missed-extractions obtained by local features alone. For ranking purpose, we classified the character strings into headings and content based on the board layout. By comparing the string using a keyword user dictionary, we assigned a high priority to the matched strings. We conducted an experiment for character-string extraction using 164 images (3933 characters). As a result, we obtained a good extraction rate of 91.4%, as compared to only 75.2% in the previous method. For the 121 images, including sign boards with more than one string, we obtained a heading-extraction rate of 55.4% and a correct-ranking rate of 67.7%

View full abstract

We have been studying two-dimensional displays using afterimages caused by saccadic eye movement. We need to investigate the spatial and temporal characters of saccade-induced afterimages, to develop a design theory on these displays. However, there have been few studies on temporal characters. We, therefore, tried to determine how long afterimages induced by saccades persisted. Our results indicate that the duration of saccade-induced afterimages last as long as that of afterimages induced by a moving illuminant with the eyes fixed, although various perceptional features are suppressed around the time of the saccade. Our results also suggest that saccade-induced afterimages can be grouped with other visual stimuli after the shape of the afterimage is formed.

View full abstract

We sometimes see a large cloud on optical disks that have been left for a long time in ordinary atmosphere. The higher the recording-density of the optical disk, the smaller the recorded size of each mark, the adhesion cannot be ignored for the drive performance. We used (NH₄) ₂SO₄ to represent the cloud and a test apparatus to investigate the relationship between the optical disk conpound and the adhesion of (NH₄) ₂SO₄. We showed the relationship between the adhesion and the read error rate, and discussed the influence of disk surface cloud on disk drive performance. Sector errors and ECC errors appeared when the adhesion was more than 2 % when the drive system used error correction code of RSPC (Inner code (182, 172, 11), Outer code (208, 192, 17)).

View full abstract

We have developed two types of real-time video watermarking systems using a VLIW-based media processor. One is an embedding system for HDTV (High-Definition TV), and the other is an embedding system for standard TV. These systems preserve the picture quality of each video frame by using motion estimation and contour detection that optimizes the strength of the watermark. We have developed a watermarking module that incorporates three MAPCAs (media accelerated processors for comsumer appliances) as art of a real-time embedding system for standard TV. For HDTV, we have developed a watermarking system with seven watermarking modules. The system divides an HDTV video into seven segments so that the seven independent modules work in parallel.

View full abstract

Relating audio-visual events is important for constructing an artificial intelligent system, which can acquire the audio-visual knowledge of moving objects through active observation without a supervisor. This paper proposes a method for relating multiple audio-visual events observed by a camera and two microphones in the scene that a moving object is not in the view of the camera. As corresponding cues, we use Gestalt's grouping law : simultaneity of sound onsets and changes in movement, similarity of repetition between sound and movement, similarity of direction between sound and movement. Based on the correlation coefficient between sound onset series and STI time series, the component of frequency at sound onset is related to the spatiotemporal invariant sequence (STI sequence) of movement. We experimented in the real environment and obtained satisfactory results showing the effectiveness of the proposed method.

View full abstract

A method retrieving videos is presented by utilizing vector quantization and latent semantic analysis. Each video is represented by a sequence of signatures through the vector quantization of frame datasets. Latent semantic analysis is then applied to the signature with a video matrix. We verified through experiments that dimensionality reduction in latent semantic analysis increases the speed and precision of retrieval. Making vector quantization more robust further improved the performance of similarity searches.

View full abstract

We previously presented a method of selectively sharpening images based on a coupled nonlinear diffusion process that was composed of a nonlinear diffusion term, a fidelity term, and an isotropic peaking term. It could sharpen blurred edges without increasing noise visibility. This paper extends our method to the removal of blurs in images due to image motion. Motion blur is not only shift-variant but also anisotropic. To adapt our method to motion de-blurring, we replaced our isotropic peaking term, with an anisotropic peaking term steered in the direction of motion. We then derived discrete calculus to adapt it to the direction of motion. Through experiments with a test image containing artificial motion-blurs, we quantitatively evaluated sharpening. Our method with the anisotropic peaking term performed better than our prototypal method with the isotropic peaking term, and was robust against errors in the direction of motion.

View full abstract