A method is presented for measuring accurate and dense 3D shapes of an object from two views captured with a moving consumer digital camera. Since existing 3D measurement systems require special and expensive measurement equipment such as laser scanners or technical knowledge such as camera calibration, it is difficult for ordinary consumers to use existing systems practically in their daily life. An easy-to-use 3D measurement system was achieved by using a Structure from Motion algorithm based on feature matching to estimate camera parameters and area-based matching to determine an accurate and dense correspondence. Experiments demonstrated that the proposed method makes it possible to measure a 3D shape with accuracy comparable to that of a laser scanner.

View full abstract

Theoretically, it is expected that a display based on an electro-holography shows three-dimensional (3D) images which satisfy all physiological factors of the depth perception of persons. A novel electro-holographic display for binocular vision was fabricated for measurements of physiological responses to 3D images. The display has a wide field of view, full parallax, and also ability to display 3D images that provide accurate accommodative and convergence stimuli correlated with depth position estimated by the viewers. The vergence and accommodation responses of viewers to the 3D images were measured objectively using an infrared optometer, and moreover depth perceptions were assessed. The comparison between measured values to the images and to real objects indicated that 3D images were displayed at the correct depth position. The results suggest that an electro-holographic display satisfy physiological factors of the depth perception.

View full abstract

Depth-image-based rendering (DIBR) is a technique for synthesizing novel views of a scene from an image and its depth map. Many DIBR methods in the literature blurred the depth map before 3D warping to make the holes become smaller. However, the PSNR values when using methods with depth map blurring are lower than those when using methods without blurring. On the other hand, conventional methods without blurring failed when the scene was a complex 3D structure. In this paper, a DIBR method without depth map blurring is proposed for synthesizing novel views by 3D warping followed by depth map and hole interpolation. To interpolate holes after 3D warping naturally, it is important to recognize the objects surrounding holes especially when the scene structure is complex. To recognize objects in the image, the input image is segmented into small regions based on color and depth. Each hole is extrapolated by extending the surrounding small regions of the hole while considering the visibility and continuity. The experimental results demonstrate the efficiency of our method.

View full abstract

[in Japanese]

View full abstract

[in Japanese]

View full abstract

Super Resolution (SR) has recently become a household phrase in Japan. Commercial products such as TV sets, BluRay players, and digital cameras are equipped with a SR function and are sold on the market. Although there are several types of SR technologies for commercial products and their details are not given to the public at all, some of them are not even new technologies and are just modified conventional ones. However, according to the official web pages of certain manufacturers of them, they are being marketed to the public as all new technologies. One of the new SR technologies of theirs is super resolution image reconstruction (SRR). Most SR technologies have complex algorithms such as iterations under the constraint condition and precise motion vector detection. Although SRR also has a complex algorithm, it is the only SR technology for real time equipment. However, in principle, there is a limitation on resolution improvement with SRR even if it works perfectly. In this paper, this limitation is discussed.

View full abstract

[in Japanese]

View full abstract