Stereoscopic images disparity for predictive coding

In order to create a 3-D system with two images (right and left) that must be transmitted simultaneously, an efficient coding appropriate for stereoscopic images must be developed to reduce the large amount of information contained in the images. As an initial step in realizing such a coding scheme, the statistical characteristics of stereoscopic images and the possibility of stereoscopic image data compression utilizing the cross correlation between the right and left images are considered. The results indicate that data compression is possible by shifting one image horizontally and subtracting it from the corresponding area of the other. Experiments to compare the capability of generating disparity by stereoscopic TV images and human stereoscopic vision have been performed. The results indicate that data compressions without 3-D image deterioration is possible by reducing high-spatial-frequency disparity above 3 cycles/degree. It is concluded that efficient image coding for stereoscopic images can be achieved by utilizing the correlation between two images and human stereopsis characteristics, making possible low-cost, more realistic 3-D visual communication