Compressed image quality non-parameter evaluation method on basis of multiscale decomposition
A multi-scale decomposition and image compression technology, applied in the field of no-reference evaluation, can solve the problems of poor quality evaluation robustness, block position shift, and unsatisfactory evaluation results, and achieve the effect of improving robustness
Inactive Publication Date: 2014-01-22
SHANGHAI UNIV
View PDF2 Cites 6 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0008] However, these methods usually assume that the block position of the image is known. When the compressed image is processed by low-pass filtering, cropping, etc., the block effect in the image is weakened or the block position is shifted, resulting in unsatisfactory evaluation results. less robust
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0043] Two preferred embodiments of the present invention are described in detail as follows:
[0044] image 3 is an original grayscale image with a size of 512×512, for the convenience of explaining how the present invention works, firstly image 3 The images shown were JPEG compressed with a compression quality factor of 60.
[0045] The present embodiment will be further described below in conjunction with the accompanying drawings.
[0046] The non-parametric evaluation method of compressed image quality based on multi-scale decomposition of the present invention, its specific steps are as follows:
[0047] (1), extract the edge pixels of the compressed image, and get all the edge pixel sets E and its corresponding set of coordinates F ;Three layers of non-subsampling contourlet transform are performed on all pixels of the compressed image, according to the corresponding coordinate set F All elements in the judgment edge pixel set E Each element in belongs to one of...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention discloses a compressed image quality non-parameter evaluation method on the basis of multiscale decomposition. The method comprises the following steps: firstly, extracting edge pixels of a compressed image and constructing an image edge gradient characteristic factor by adopting non-subsample contourlet transform; secondly, carrying out discrete cosine transform (DCT) on the compressed image to obtain an image smooth characteristic factor; and finally, fusing the image edge gradient characteristic factor and the image smooth characteristic factor to obtain an evaluation value formula of compressed image quality non-parameter evaluation so as to implement non-parameter evaluation on quality of the compressed image. According to the invention, the image edge gradient characteristic factor and the image smooth characteristic factor are fused to obtain an evaluation value of compressed image quality non-parameter evaluation and structural distortion and smooth distortion which are caused by the compressed image are comprehensively reflected, and thus, the quality of the compressed image can be effectively evaluated; and when the image is subjected to low-pass filtering or cut down, the distortion conditions of the image on the aspects of structural distortion and smooth distortion still can be effectively reflected, and thus, robustness of image quality evaluation is also improved.
Description
technical field [0001] The present invention relates to a method for decomposing digital image content for reference-free assessment of compressed image quality using signal processing and computer technology. Background technique [0002] Since compressed images are widely used, evaluating the quality of compressed images is helpful to analyze the quality of images and evaluate the performance of various compression algorithms. In quality assessment, according to whether there is an original image as a reference, it is usually divided into parameterized evaluation and non-parametric evaluation. Compared with parametric quality assessment, non-parametric quality assessment of digital images does not require the original uncompressed image, so it has better application value. [0003] The common compressed images mostly use the block method, and each block is compressed separately, which to a certain extent causes a certain difference between the image blocks and the adjacen...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Applications(China)
IPC IPC(8): H04N17/00H04N19/625
Inventor 钱振兴王文文张新鹏
Owner SHANGHAI UNIV