Region image coding techniques

Region and texture coding is a second generation technique that uses features identified in the original picture to decompose it into two components which are then coded separately. The purpose of taking this approach is that by finding and encoding the perceptually important features of the image they will be preserved. By separating the signal into two components of very different character it should be possible to design efficient coders for each of them. The paper is a comparative study of three techniques that can be employed to code the region signal. The authors detail how the region signal is generated and three algorithms which can be used to code it. A comparative experiment is performed and the results presented.<>