Two-dimensional Discrete Wavelet Transform (2-D DWT) is an effective tool for
image compression and hence attracts much attention in recent years. In this paper, we propose
an adaptive quantization approach which quantizes the DWT coefficients according to their
characteristics. Unlike the uniform quantization which uses a fixed quantization interval, the
proposed scheme applies two asymmetric interval widths to one single sub-band. According to
the experiment results, it leads to a superior image quality while maintaining the same
compression ratio.