Abstract

This paper presents a novel hardware-oriented image compression algorithm and its very large-scale integration (VLSI) implementation for wireless sensor networks. The proposed novel image compression algorithm consists of a fuzzy decision, block partition, digital halftoning, and block truncation coding (BTC) techniques. A novel variable-size block partition technique was used in the proposed algorithm to improve image quality and compression performance. In addition, eight different types of blocks were encoded by Huffman coding according to probability to increase the compression ratio further. In order to achieve the low-cost and low-power characteristics, a novel iteration-based BTC training module was created to get representative levels and meet the requirement of wireless sensor networks. A prediction and a modified Golomb-Rice coding modules were designed to encode the information of representative levels to achieve higher compression performance.

Description

.