|M.Sc Student||Roterman Yalon|
|Subject||Color Image Coding Using Regional Correlation of|
|Department||Department of Electrical Engineering||Supervisor||Professor Moshe Porat|
Natural scene images are characterized by high correlation between their RGB components. Many color compression techniques reduce the redundancies between the RGB colors, by transforming the colors primaries into a decorrelated color space such as YIQ. As the human visual system is more sensitive to details in luminance than to details in chrominance, the color information can be compressed at a higher rate. The JPEG compression algorithm is a typical example of this approach where the I and Q information is encoded separately at lower rates than the Y.
The high regional correlation between the RGB components of color images is used as a basis for a new coding technique. The correlation between the color primaries suggests a localized functional relation between the components. In this work we approximate subordinate colors in smooth segmented regions of an image as a function of one of its base colors. Unlike JPEG, which arbitrarily sub-divides the image into 8x8 blocks, it seems more efficient to divide the image according to its natural boundaries. This is carried out in the proposed algorithm using the ‘region growing’ method.
The encoded results are superior to those obtained by other compression techniques that are based on decorrelated color spaces, as in JPEG. We also introduce an option of progressive transmission, which could be used for slower communication channels.
Our conclusion is that the use of correlation in color images compression could be superior to the traditional decorrelation techniques, in particular for images with high localized correlation.