SAR image denoising method based on AD-NSCT algorithm
An image and algorithm technology, applied in the field of SAR image denoising, can solve the problems of not making full use of image redundant information, denoising results not reaching smoothness and edge preservation, etc., achieving good pixel spatial distribution characteristics and improving visual effects , the effect of strong inhibitory ability
Active Publication Date: 2013-11-27
XIDIAN UNIV
View PDF3 Cites 9 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
However, this data distribution model is only a reasonable assumption, and does not make full use of the redundant information of the image. The method of calculating the likelihood ratio needs to be improved, and the denoising result does not achieve the best smoothness and edge preservation.
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
[0028] refer to figure 1 , the specific implementation process of the present invention is as follows:
[0029] Step 1. Input a SAR image and perform non-subsampling Contourlet transformation on it:
[0030] Set the number of decomposition levels to n levels =[2222], using the non-subsampling Contourlet transform of the pyramid filter and the directional filter bank to decompose the input SAR image; that is, the input SAR image is non-subsampled according to the subbands of four scales and four directions Contourlet transform, generate coefficient matrix c (l，d) , where l represents the scale and d represents the direction subband.
[0031] Step 2. For the coefficient matrix c (l，d) , calculate the noise standard deviation σ of each scale and each direction subband (l，d) :
[0032] σ (l，d) =Median(|c (l，d) -Median(c (l，d) )|) / 0.6745,
[0033] Among them, Median() is the function to find the median value, c (l，d) is the coefficient matrix, l ∈ {1, 2, 3, 4}, d ∈ {1, 2...
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 an SAR image denoising method based on an AD-NSCT algorithm. The phenomenon that an existing image denoising method is blurred in edge and not thorough in noise removal is mainly improved. The implementation process of the method comprises the steps that (1) an SAR image is input; (2) non-subsample contourlet conversion is carried out on the SAR image according to sub-bands of four scales and in four directions to generate a coefficient matrix; (3) the noise standard deviation of the coefficient matrix is calculated; (4) the covering film matrix of the image is calculated; (5) the Affinity matrix of the image is calculated, and an Alpha matrix is calculated by combining the coefficient matrix; (6) the corresponding covering film matrix is corrected through the Alpha matrix, and a posteriori ratio is calculated; (7) a priori ratio is calculated through the corrected covering film matrix; (8) the zoom factor of the image is calculated through the posteriori ratio and the priori ratio; (9) zoom processing is carried out on the coefficient matrix, inverse transformation is carried out on the result, and the denoising of the SAR image is achieved. According to the SAR image denoising method based on the AD-NSCT algorithm, high smoothness is ensured, meanwhile, the edge detail information of the SAR image can be well maintained well, and the vision effect of the denoised SAR image is obviously improved.
Description
technical field [0001] The invention belongs to the technical field of image processing, relates to a SAR image denoising method, can be applied to the fields of image processing and recognition preprocessing, and is the basis for subsequent higher-level analysis and processing of images. Background technique [0002] Due to the inherent speckle noise in the Synthetic Aperture Rader (SAR) image, it is not conducive to the automatic analysis of the scene in the image and the understanding of the SAR image. Therefore, the removal of speckle noise is very important to the subsequent processing of SAR images such as edge detection and image segmentation. Compared with optical imaging, SAR has its unique advantages, that is, it is not limited by time and clouds. As a landform and surface object sensor, SAR covers a wide range of areas. Its advantages of high resolution, large area and all-weather work make it widely used in the fields of earth remote sensing, marine research, re...
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 IPC IPC(8): G06T5/00
Inventor 田小林焦李成郭开武王爽张龙马文萍马晶晶
Owner XIDIAN UNIV