Method, system and apparatus generating up and down sampling filter

Abstract

The invention provides a method for generating an up sampling filter and a down sampling filter, which comprises: processing down sampling and up sampling on the video signals fitting a markov mode to attain reconstruction signals; calculating out the energy of the reconstruction signals; according to the biorthogonal wavelet theory carrying parameters and the energy of the reconstruction signals, calculating out the parameters of the wavelet filter corresponding to the maximum energy of the reconstruction signals, to generate an up sampling filter and a down sampling filter. The invention further provides a method for adopting an up sampling filter and a down sampling filter to realize spatial graded video code, a device and a system. The invention can minimize the coding energy loss, to improve the quality and performance of video signal coding.

Description

Method, system and device for generating up-down sampling filter and implementing coding technical field The invention relates to coding technology, in particular to a method, system and device for generating up-down sampling filters and realizing coding. Background technique With the development of Internet, wireless communication and other technologies, the demand for transmission of video-based multimedia data is also increasing rapidly. If all multimedia data is transmitted without compression, it needs to occupy a large bandwidth and a storage medium with a large capacity, so the transmission volume of multimedia data is often very large, and it is also very difficult to implement, so it is necessary to compress the multimedia data first. Encoded before transmission. In addition, in the transmission process, the encoding and transmission are required to have the ability to adapt to network bandwidth over time, anti-error performance, and the ability to support spatia...

AI Technical Summary

Problems solved by technology

In fact, in practical applications, the up-down sampling process using the above-mentioned up-sampling filter and down-sampling filter is independent and lacks correlation. Therefore, the video signal encoded by using these filters will appear distorted, fast Effects, etc., resulting in poor visual effects for the audience, so the encoding quality of the video signal cannot be guaranteed

In addition, because the up-sampling filter and down-sampling filter are too long, the computational complexity is too high, resulting in low encoding speed

In order to solve the above-mentioned problems that the above-mentioned up-sampling filter and down-sampling filter are too long and the encoding speed is too low, JVT-U147 has improved the above-mentioned up-sampling filter and down-sampling filter, and proposed the following up-sampling filter and down-sampling filter Filter scheme: h5=[-1262-1] / 8; h7=[-10916 90-1] / 16

Due to shortening the length of the filter, the speed of encoding using this filter has been improved to a certain extent, but the up-sampling and down-sampling processes of the up-sampling filter and the down-sampling filter are also independent and lack of correlation. , therefore, the use of these filters in the encoding process cannot guarantee the encoding quality of the video signal

It can be seen that in the process of spatially scalable coding supporting the H.264SVC coding standard, the filtering performance of the up-sampling filter and down-sampling filter generated by using the existing technical scheme is poor. In addition, using the existing up-sampling filter and down-sampling filter Filters encode poor quality

Images