国際標準化団体MPEGでは, 3DAVという活動名で3D映像の標準化が検討されている. 検討されている3D映像は, 実写を基にした3Dシーンに対して視点位置や視線方向を変更できるインタラクティブな機能を有する. CGとは異なり自然画像を使用するため, よりリアルな映像を再生できる. 本稿ではMPEGで検討されている3DAV活動の概要と歴史, ならびに現在進行中の活動内容を紹介する.

抄録全体を表示

We developed a communication system using internet protocol technology for broadcasting stations. First, networks are established between a private branch exchange (PBX) server and areas such as studios, sub-control rooms, local broadcasting stations, and the incident scenes where communication systems are used. The PBX server returns the audio signal to each user by mixing other audio signals for each group and produces arbitrary conversation groups for members who want to talk. Access points are established in the studios and sub-control rooms, so that wireless LAN communication is available. Each terminal can be connected to a headset (neck-strap-type earphone with a microphone) and can be supplied with power from an external battery pack, which runs for approximately eight hours. A connection from outside wires (city telephones) or extensions to the PBX server via a gateway can combine two different functions of existing communication systems, intercom and telephone for coordination. An existing intercom system can operate together with the new communication system if both systems are connected via a gateway.

抄録全体を表示

We developed a display for presenting 3-D shapes using a pin-rod matrix. Due to a wide range of movement, the display is capable of presenting large-scale, dense objects such as human faces or geographical features. We used a coil-type shape memory alloy as a pin-rod actuator for building our display. We first designed and evaluated the system components. We then constructed three different prototypes. The first prototype has a 1x64-pin-rod matrix, and each of the other two has a 16x16-pin-rod matrix. The prototypes have a 30-mm range of motion and a 5-mm pin interval. The prototypes are sufficiently accurate for presenting subtle surface structures. Our goal is developing a dynamic display that moves in real-time and that also works as both an output and input interface.

抄録全体を表示

Pay-TV, which allows viewers to select the programs they want to watch, is commonly available. A viewer wants to buy a desired program after he/she understands its content. A broadcaster wants to protect a program's content and advertize during the program. Existing pay-TV service is not able to meet both requirements. We developed a method for embedding visible watermarks that satisfies both viewers and broadcasters on pay-TV. The visible watermark is embedded in a program for copyright protection and advertizing purposes. In addition, the logo of the watermark is split into two keys. The broadcaster has one key, and the other key is embedded in unnoticeable regions of the program. The broadcaster is thus able to remove the visible watermark by using the key. Furthermore, the method is resistant to attack by the removal of the visible watermark because it is changed by image features temporally. Experimental results show that understanding the content of a program is possible from a sample image. In addition, the image that is produced is comparable to the original image.

抄録全体を表示

With the digitization of broadcasting services and the expansion of broadband infrastructure in recent years, high-quality video-broadcasting service is beginning to be provided over communication media. Furthermore, having an attractive service in which broadcasting and communication cooperate is desirable. We describe a technology and a system for the integrated management and control of various kinds of content distribution, such as real-time broadcasting, broadcasting based on home servers, video on demand, and file downloading, by using the metadata that describes the service control conditions and the content description metadata that conforms to the TV-Anytime Forum specifications.

抄録全体を表示

Natural 3-D images can be generated by displaying high-density directional images. The 3-D camera technique that generates high-density directional images can be implemented by interpolating images that were captured by a horizontally-aligned camera array using image-based rendering techniques. Three cameras in the camera array that was located in front of an object were used to obtain right- and left-side depth maps of the object. Using these depth maps in the image interpolation process enabled us to fix the camera pitch independently of the depth of the object. We found that when numerous cameras were aligned horizontally and the cameras in front of the object were selectively used, movements of the object in the horizontal and depth directions were allowed because the camera pitch did not have to be changed. The preliminary experimental results indicated that 64 directional images were generated from eight captured images.

抄録全体を表示

This paper introduces a preprocessing method for Free Viewpoint Television (FTV). In every multi-camera system, it is difficult to precisely align cameras in the desired place. For FTV, this misalignment reduces the efficiency of image compression and the quality of interpolated images. We propose a preprocessing method that transforms the captured multi-camera images by using projective transformation matrices to compensate for this misalignment. This transformation is able to improve them because the multi-camera images become more correlative. Furthermore, we introduce a method for automatically finding the projective transformation matrices using the epipolar geometry. Experimental results show that PSNR gains of up to 3.0 dB in comparison with compressions without any processing, when MPEG-2 is used under conditions ranging from 1.6 to 3.0 Mbps. The quality of the interpolated images is also apparently improved.

抄録全体を表示

Detecting skin regions of an image is one of under interest areas in computer vision and graphics. It can be primary step in several applications like advanced human-computer interaction, biometric authentication or contextual image retrieval. Several different researches and challenges have been done to classify the image regions into two groups of skin and non-skin areas. Most of them discussed well in theoretical aspects like effect of different color spaces or classifiers on large datasets. Here it is tried to have practical look to the problem. In a real final system we will have some limitations on speed, memory and even training dataset in different conditions. After a short overview to some common previous methods, a novel skin model is proposed based on LVQ neural networks. Consequently a study on homogenization methods gives more accurate choice for this step. It is shown with the experiments that this scheme can reserve advantages of several different methods at once. It can perform comparable accuracy of 87.92% with acceptable speed while covering wide variety of skin color with a limited size of training data.

抄録全体を表示

We have developed an extremely high-resolution display (Super Hi-Vision display) system that has quadruple the amount of the horizontal and vertical resolution of HDTV. The extremely high-resolution display consists of two LCD projectors, for green and another for red and blue, so convergence errors between green and red/blue images exist on the screen. We introduce a method for automatically adjusting the convergence that detects convergence errors by displaying mountain-like patterns using a digital-camera. The projector for green images has a resolution equivalent to 7680×4320 pixels with a spatial pixel-offset method using two 3840×2160 LCD panels. In this case, two green panels must be located at an exact spatial offset of 0.5 pixel. We developed a method for detecting a misalignment of spatial offset by using two sine-wave patterns with different phases.

抄録全体を表示

A model proposed in this paper represents the hole process of product-innovation-type technology development. the most important element in this model is the Entrepreneur-ship with individuals or their teams, which promote disruptive innovation. At this point, this model is different from MOT which has recently attracted considerable attention in industrial and educational circles. Specifically this model explains how “Techno-entrepreneurs”, who lead projects, find out future possibility in the uncertainty of technology. Usually in disruptive innovation several interactive feedback loops are needed among research, development, and sales, so the model represents all these interaction. Furthermore case for verifying this model's appropriateness, a history of one disruptive innovation is described from its technology through its grow stage.

抄録全体を表示

Input interface devices worn by users are desirable for wearable computers. A tiny wristwatch-like TV camera was worn on the wrist of one hand and a retro-reflective maker was worn on a finger of the other hand. They were used for capturing hand gestures as a wearable input device. While looking at the positions of the camera and their own hands, the users pointed at captured areas with the marker for inputs by visual estimation. Near-infrared illumination, a retro-reflective optical marker, and a visual light-cut optical filter were used for extracting the marker from the captured images, and the position of the marker was calculated based on the captured images. LEDs mounted on the camera were also used for indicating the position of the maker. the pointing performances were empirically measured under various indicator conditions. Based on the experimental results, the device is sufficient for pointing at nine areas of a captured area with a high degree of accuracy.

抄録全体を表示