Adaptive matrix design for LDPC based image processing system

In the field of communication, coding the data is the most crucial part. Since transmission channel is very much susceptible to noise, it becomes imperative to channel code the data to be transmitted. This paper presents an improved and faster Low Density Parity Check (LDPC) based algorithm, for channel coding grayscale images of ´n*n´ dimension. It constructs the encoded image, by arranging bits into 8n data vectors and modulating it with BPSK modulation technique. The novel approch for designing an adaptive matrix as a function of input image size is described in the paper. The decoding algorithm is well equipped with a probabilistic approach. Further analyzing the performance of the algorithm shows improved restitution of images, that are affected by AWGN noise. Hence, LDPC codes prove their dominance over other coding schemes viz. Turbo codes, BCH etc, for applications such as medical imaging, deep-space communications and multimedia.