Digital Image Compression Using Adaptive Btc Method Based On Weber – Fechner’s Measure

Abstract

Using Weber-Fechner's law in digital image compression is an efficient method for decreasing computational complexity and improving compression results in low bit rates with the same reconstructed image quality for conventional coding method (without using Weber-Fechner's condition). Our purpose here is to increase the compression ratios (low bit rates) by using a new coding system based on the condition of Weber- Fechner's law which is proposed to classify low, constant and high detail blocks in an image, then low or constant detail blocks are compressed by only sending the mean of the original image block, while the high detail blocks are compressed using AMBTC or AMBTC+VQ system, then send the two reconstruction values and the bit map when we use the AMBTC method but when we use the AMBTC+VQ methods, we send the two reconstruction values and the index of the codeword. As documented in experimental results, using Weber-Fechner's condition in AMBTC or AMBTC+VQ the bit rate and computational complexity are much lower than using the conventional for AMBTC or AMBTC+VQ methods with the same image quality.