Quality evaluation method for video lossy compression

Abstract

The present invention provides a method for evaluating lossy compression quality of image. The method comprises the following steps: selecting related image as testing image as testing image according to a set image selecting principle; executing data compression to the tested image with an image compression method to be evaluated thereby obtaining compression image; decoding for obtaining reconstructed image; analyzing the tested image and reconstructed image through a plurality of image evaluation aspects thereby evaluating the adopted image compression method; wherein, the plurality of image evaluation aspects can execute full evaluation to the compressed image quality, and simultaneously the plurality of image evaluation aspects have no relativity. The method can execute full evaluation to the quality of compressed image, is suitable for the original image with different landform categories and texture abundance degrees, and can provide a reference for the improved design of spaceborne compression algorithm developed by our country.

Description

technical field [0001] The present invention generally relates to image lossy compression quality evaluation methods, and more specifically, relates to a combination of visual cognition methods, digital image analysis methods, and digital photogrammetry techniques for interpretation quality evaluation, geometric quality evaluation, and application analysis evaluation. Image lossy compression quality evaluation method. Background technique [0002] Image compression techniques are widely used in various fields, for example, in the field of remote sensing, various image compression techniques are often used. [0003] my country's first high-resolution optical stereo-mapping satellite, Ziyuan-3, is planned to be launched around 2011. Ziyuan No. 3 satellite adopts the adaptively improved Ziyuan No. 2 satellite platform, equipped with four cameras, including two front-view and rear-view panchromatic TDI CCD cameras with a ground resolution better than 4.0 meters; one ground reso...

AI Technical Summary

Problems solved by technology

[0017] 3) In most cases, due to the limitation of sensor performance, the original image itself before compression is an incomplete description of the real scene

[0024] 1) Redundancy between pixels;

[0025] 2) Psychovisual redundancy;

[0027] 4) Structural redundancy;

[0029] 6) Temporal redundancy

[0041] (3) Gray standard deviation

[0083] (2) The impact of data compression on different terr

Images