Fast and Efficient 4x4 Tchebichef Moment Image Compression

Orthogonal moment functions have long been used in image analysis. This paper proposes a novel approach based on 4x4 discrete orthogonal Tchebichef moment for fast and efficient image compression. The method incorporates a simplified mathematical framework technique using matrices as well as a block-wise reconstruction technique to eliminate possible occurrences of numerical instabilities at higher moment orders. Then the 4x4 Tchebichef Moment Transform and Discrete Cosine Transform have been compared. The results show that the 4x4 Tchebichef moment has significant advantages over the other technique in terms of its error reconstruction, average bit-length of Huffman codes and image quality. Moreover, Tchebichef moment provides a compact support to sub-block reconstruction for image compression. Tchebichef Moment Compression clearly performs potentially better for broader domains on real digital images and graphically generated images.