Improving Image Reconstruction Accuracy Using Discrete Orthonormal Moments
Several pattern recognition applications use orthogonal moments to capture independent shape characteristics of an image, with minimum amount of information redundancy in a feature set. Legendre, Zernike, and Pseudo-Zernike moments are examples of such orthogonal feature descriptors. An image can also be reconstructed from a sufficiently large number of orthogonal moments. Discrete orthogonal moments provide a more accurate description of image features by evaluating the moment components directly in the image coordinate space. This paper examines some of the problems associated with the computation of large order Tchebichef moments, and proposes an orthonormal version to improve the quality of reconstructed images.