Image Data Compression by Adaptive Thresholding of Wavelet Coefficients

Image compression is performed to reduce the bit rate for image transmission and the required storage in the case of image archival. Typically, it involves a transformation of the image to reduce the correlation and redundancy between pixels. The wavelet transform has gained wide acceptance as a tool to decorrelate most image sources. The transformed image is decomposed at different scales of details and approximation. Such decomposition facilitates image compression and progressive transmission. In this study, a system and method for image compression is presented based on the wavelet transformation. A threshold is determined adaptively based on the desired compression ratio and only significant coefficients are retained.