An image compression method based on orthogonal wavelet packet transform

The main reason why the image data can be compressed is that generally, the original image data is highly correlated and it contains plenty of redundant information. The purpose of image compression encoding is to erase the redundancy, represent and rebuild the image with as a small number of bits as possible with the given distortion in order to make it consistent with the requirements of the applications. By studying the nature and coefficient feature of each sub-band of the image after wavelet decomposition, this paper proposes an image compression method based on orthogonal wavelet packet transform. In other words, the mean image signal of the high-frequency components of the image is divided into four frequency bands after one wavelet transform, namely the high-frequency parts in the horizontal direction, the vertical direction and the diagonal direction and the low-frequency part, which will continue to decompose. In this way, the image signal has been decomposed to the sub-image signals different spatial resolutions, frequency characteristics and direction characteristics so that simultaneous processing can be achieved to long-term features of low frequency and short-term features of high frequency. In fact, orthogonal wavelet packet transform amounts to a low-pass filter. The main part an image presents itself is its low-frequency part while most of the high-frequency part is close to 0. The image compression based on orthogonal wavelet packet transform is to use its low-pass features to filter the highfrequency part and preserve the low-frequency part. The simulation experiment proves that the algorithm of this paper effectively overcomes the limitations existing in the complicated image compression in order to make the decomposition of the image signal more consistent with the visual characteristics of humans and the requirements of data compression.