Improved SPIHT image coding and decoding method based on energy leakage and amplification

Abstract

The invention discloses an improved SPIHT image coding and decoding method based on energy leakage and amplification. Firstly, a secondary wavelet transform is applied along the main gradient direction of the highest frequency subband to further concentrate the energy of the high amplitude coefficients with amplitude correlation on a small number of coefficients and increase the height of the spatial direction tree. Secondly, the high frequency subbands based on wavelet transform will retain a small amount of medium, the Canny operator is used to detect the main edges and contours of the decoded image, and a high-pass filter with smooth falling edge is used to extract and amplify the residual middle and low-frequency information in the high-pass coefficients iteratively, so as to enhance the decoding effect at the main edges and contours. The experimental results show that the invention can effectively improve the subjective and objective quality of the decoded image under the low coderate.

Description

technical field [0001] The invention relates to the field of image processing, in particular to an improved SPIHT image encoding and decoding method based on energy leakage and amplification. Background technique [0002] With the continuous development of network multimedia technology and its applications, people's requirements for image coding continue to increase, which not only requires high reconstruction quality of decoded images, but also requires it to be able to adapt to application requirements such as network heterogeneity and terminal computing capabilities. . This requires a good solution to remove all kinds of redundancy under the premise of ensuring image quality, use as few bits as possible to represent image information, and at the same time make the image encoding stream well adapt to fluctuations in network bandwidth And the decoding requirements of terminals with different computing capabilities. Scalability of the coded stream of an image refers to the...

AI Technical Summary

Problems solved by technology

[0006] However, most of the existing methods ignore the fact that wavelet transform is not the optimal sparse representation tool for 2D images. After the image undergoes wavelet transform, the high-frequency subbands still contain Large number of wavelet coefficients with large magnitudes, unable to achieve optimal approximation and energy concentration of complex edge singularities

In this case, two problems inevitably arise: on the one hand, the low frequency to high The mechanism of frequency layer-by-layer processing makes it impossible to effectively decode high-frequency coefficients at low bit rates, resulting in serious information loss, and the reconstructed image produces obvious "ringing" effects at the edge contours

On the other hand, in the filtering process of wavelet transform, the low-pass filter and high-pass filter used are not ideal filters. The high-frequency energy below the cut-off frequency will leak to the low-frequency sub-band, and the low-frequency energy above the cut-off frequency Energy also leaks into the high frequency subbands

Images