The rate-distortion optimized quantization algorithm in Compressive Sensing

Abstract

Rate-distortion optimization greatly improves performance of compression coding system. In this paper, the rate-distortion optimized quantization algorithm is proposed for block-based Compressive Sensing. Compressive Sensing is the emerging technology which can encode a signal into a small number of incoherent linear measurements and reconstruct the entire signal from relatively few measurements. In the algorithm the sampling measurements are quantized optimally based on the rate-distortion theory. For the coefficients near dead-zone the quantization level with best rate-distortion performance is chosen. Moreover, in order to acquire the best performance, a fast Lagrange multiplier solving method is proposed to find the optimal slope λ* of the rate-distortion curve at the given bit budget. Experimental results show that the proposed algorithm improves objective and subjective performances substantially. The average gain about 0.7dB can be achieved with the same rate.