Convex projections restoration of Hadamard naturalness-preserving transform coded images

This paper investigates an image coding system using the Hadamard naturalness-preserving transform (HNPT). It focuses both on the encoding and the decoding processes. Symmetric Lloyd-Max quantizers are designed to encode the HNPT coefficients. The theory of projections onto convex sets (POCS) is used as a basis for the design of an iterative reconstruction algorithm. A constraint set is defined that imposes consistency with known values. Experimental results to date demonstrate the convergence of the restoration algorithm.