CN103077499B

CN103077499B - SAR (Synthetic Aperture Radar) image pre-processing method based on similar block

Info

Publication number: CN103077499B
Application number: CN201310008103.8A
Authority: CN
Inventors: 钟桦; 焦李成; 于艳青; 马晶晶; 马文萍; 侯彪; 黄捷; 张小华
Original assignee: Xidian University
Current assignee: Xidian University
Priority date: 2013-01-09
Filing date: 2013-01-09
Publication date: 2015-04-08
Anticipated expiration: 2033-01-09
Also published as: CN103077499A

Abstract

The invention discloses an SAR (Synthetic Aperture Radar) image pre-processing method based on a similar block. The problem that image texture details and compressed strong reflecting point targets cannot be kept by the current method is mainly solved. The realization process comprises the following steps of (1) inputting an SAR image, and calculating a variance factor matrix of the image; (2) dividing an image pixel point into a smooth area and a texture detail area by using a variance coefficient; (3) performing mean filtering on the smooth area to obtain a processed pixel value; (4) performing spiral operation on the texture detail area; (5) obtaining a similarity measurement formula according to specific value probability distribution, and calculating the similarity between a to-be-processed point and eight neighbourhood pixel points; (6) taking the average value of the to-be-processed point and three points which are mostly similar to the to-be-processed point as the gray value of the pixel point; and (7) calculating the recovering value of all the pixel points to obtain a pre-processed image. The texture and the structure information of the image can be better protected while noise is suppressed, strong reflecting point targets can be well kept, and the method can be used for pre-processing the image before the image is applied.

Description

SAR image preprocessing method based on similar blocks

Technical Field

The invention belongs to the technical field of image processing, and relates to an SAR image preprocessing method based on homogeneous similar blocks, which can be used for preprocessing SAR images before application.

Background

The image formed by the synthetic aperture radar SAR has the characteristics of all weather, all time, high resolution, strong penetrating power and the like, so that the image is widely applied to the fields of target identification, transformation detection, water surface monitoring and the like. However, SAR images are prone to corruption by multiplicative noise from the continuous interference of backscatter radar reflections, and speckle noise that destroys the resolution of SAR image radiometry while impacting the performance and understanding task of background analysis.

With the widespread application of SAR images in military and civil applications, SAR image processing becomes another research focus of SAR technology. The SAR image contains a large number of coherent speckles caused by the random interference of coherent echoes of scattering points on an imaging interface, and the speckle noises can greatly reduce the effectiveness of image segmentation, edge detection, feature extraction, target identification and other information processing technologies, so that the SAR image processing cannot reach an expected target. Therefore, preprocessing before processing the SAR image is an indispensable process.

The purpose of the preprocessing is to suppress image noise while preserving feature information of the image, such as image texture, edges, and point objects. The method is greatly different from the image denoising method, which is to remove noise as much as possible, so that the texture of the image is smoothed, and the image structure information such as edges, linear bodies, points and other targets are blurred or filtered to a certain extent, which is not beneficial to the processing of subsequent images. Currently, there are two main methods for preprocessing an SAR image: 1)3 × 3 block mean method; 2)3 × 3MMSE method. The preprocessed image obtained by the 3 x 3 block mean method has better homogeneous region smoothing capacity, but the brightness of the strong reflection point target is seriously compressed, and the edge is also blurred to a great extent; the 3 × 3MMSE method can better maintain the homogeneous region smoothing ability, and also can compress the brightness of the strong reflection point target, so that the texture details of the image become fuzzy, and the subsequent processing is not facilitated.

Disclosure of Invention

The invention aims to provide an SAR image preprocessing method based on homogeneous similar blocks aiming at the defects of the prior art, so that the processed image can well keep the smooth capability of a homogeneous region, meanwhile, the strong reflection point target and the image texture structure information of the image are protected, and the subsequent processing of the image is facilitated.

The technical key point for realizing the purpose of the invention is that when the similarity is calculated by using the blocks, the image blocks are rotated, so that the similarity calculation of the similar blocks is more accurate; meanwhile, ratio distribution probability is introduced on the basis of homogeneous similar blocks, and the technical scheme comprises the following steps:

(1) calculating the variance coefficient CV of all pixel points for the L-view SAR image v with the input size of (m, n) to obtain a variance coefficient matrix K;

(2) setting a variance coefficient classification threshold T_cvClassifying the input SAR image v, if the pixel point x in the image v_i,jThe coefficient of variance in the coefficient of variance matrix K is less than a threshold T_cvIf not, executing the step (3), otherwise, executing the step (4);

(3) will be defined by the pixel point x_i,jTaking the average value of the pixels in the 3 x 3 blocks as the center as the pixel value of the pixel point after preprocessing;

(4) for pixel point x_i,j8 neighborhood pixel point x_lL =1,2, …,8, taken as pixel point x_lCentered 3 x 3 block v_lTo block v_lBy rotating to make pixel point x_i,jCentered 3 x 3 block v_i,jAnd with pixel point x_lCentered 3 x 3 block v_lThe medium homogeneous region is at the same position, v_lThe rotated block is marked as；

(5) Calculating a pixel point x_i,jAnd 8 neighborhood pixel points x thereof_lL =1,2, …,8 similarity distance based on block ratio probability:

5a) taking pixel x_i,jCentered 3 x 3 block v_i,jObtaining the neighborhood pixel point x in the step (4)_lL =1,2, …,8 flipped 3 × 3 block；

5b) Calculating the two pixel blocks v_i,jAndratio r of_i,k：

r_{i, k} = \min {\frac{v_{i, k}}{v_{l, k}}, \frac{v_{l, k}}{v_{i, k}}},

r_i,k∈[0,1]，l=1,2,…,8，k=1,2,…,9，

Wherein v is_i,kRepresenting a pixel point x to be processed_i,jCentered 3 x 3 block v_i,jThe gray value v of the kth pixel point_l,kRepresenting a pixel point x_lCentered 3 x 3 block v_lRotated blockThe gray value of the kth pixel point;

5c) calculating pixel point x by using ratio distribution probability formula_i,jRatio r of_i,kProbability of occurrence p (r)_i,k)；

5d) Defining a pixel point x to be processed_i,jAnd 8 neighborhood pixel points x thereof_lThe distance d of similarity therebetween_lComprises the following steps:

k=1,2,…,9，l=1,2,…,8；

(6) for the similarity distance d obtained in the step (5)₁,d₂,…,d₈Sorted in ascending order, the sorted result isGet the distance after the sequence asAs pixel point x_i,jLike the pixel point x_i,jAnd the mean value of the gray values of the 3 similar points is taken as a pixel point x_i,jThe preprocessed gray value.

(7) And (6) repeating the steps (2) to (6), and calculating the recovery values of all the pixel points to obtain the preprocessed image.

Compared with the prior art, the invention has the following advantages:

1. the invention adopts variance coefficient to divide the SAR image into two areas: the homogeneous region and the texture region adopt different methods for different regions, thereby improving the processing precision.

2. According to the method, the rotation operation is adopted, so that the homogeneous regions in the blocks are located at the same position, the calculated similarity is more accurate, the current point can be repaired by using the pixel points with higher similarity, and the texture and detail information of the image can be better maintained;

3. the method adopts the similarity distance based on the direct distribution probability, and can more accurately calculate the similarity between the SAR image pixel points;

drawings

FIG. 1 is a flow chart of the present invention;

FIG. 2 is a test image used by the present invention;

FIG. 3 is a graph of the results of the pre-processing of FIG. 2 using the prior art MMSE approach;

FIG. 4 is a graph of the results of a prior art averaging method for preprocessing FIG. 2;

FIG. 5 is a graph showing the results of the pretreatment of FIG. 2 by the method of the present invention

Fig. 6 is an enlarged detail view of a result graph of preprocessing of fig. 2 by the existing MMSE method.

FIG. 7 is an enlarged detail view of a graph of the results of the preprocessing of FIG. 2 using a prior art averaging method;

FIG. 8 is an enlarged detail view of the result of the pretreatment of FIG. 2 by the process of the present invention.

Detailed Description

Referring to fig. 1, the present invention comprises the steps of:

step 1, calculating the variance coefficient CV of all pixel points for the L-view SAR image v with the input size of (m, n) to obtain a variance coefficient matrix K.

1.1) calculating the Pixel Point x_i,j，i∈[1,m],j∈[1,n]Coefficient of variance CV of_i,j：

Wherein,is a pixel point x_i,jThe standard deviation of the gray values of all the pixel points in the 7 x 7 neighborhood of the center,is a pixel point x_i,jThe mean value of gray values of all pixel points in a 7 multiplied by 7 neighborhood which is taken as a center;

1.2) calculating the variance coefficient of each pixel point in the SAR image v to obtain a variance coefficient matrix K₀

K₀={CV_i,j},i∈[1,m],j∈[1,n]

1.3) to the calculated coefficient of variance matrix K₀And carrying out 3 × 3 mean filtering to obtain a filtered variance coefficient matrix K.

Step 2, setting a classification threshold value T_cvAnd classifying the input SAR image v.

2.1) setting threshold T according to input L-view image type_cv：

If the input is an L-view amplitude SAR image, the threshold value T is_cvThe following settings are set:

if the input is the L apparent intensity SAR image, the threshold value T_cvThe following settings are set:

T_{cv} = 0.82 \sqrt{\frac{1}{L}},

2.2) grouping pixels x_i,jCoefficient of variance CV of_i,jAnd a threshold value T_cvBy comparison, the images are thus divided into two classes of processing, if the coefficient of variance CV_i,jLess than threshold T_cvExecuting step 3, otherwise executing step 4;

step 3, pixel point x_i,jThe mean value of the gray values of the pixels in the 3 x 3 block as the center is used as the pixel x_i,jThe preprocessed gray value.

Step 4, for pixel point x_i,j8 neighborhood pixel point x_lL =1,2, …,8, taken as pixel point x_lCentered 3 x 3 block v_lTo block v_lBy rotating to make pixel point x_i,jCentered 3 x 3 block v_i,jAnd with pixel point x_lCentered 3 x 3 block v_lThe medium homogeneous region is at the same position, v_lThe rotated block is marked as。

Step 5, calculating pixel point x_i,jAnd 8 neighborhood pixel points x thereof_lSimilarity distance of (d):

5.1) taking pixel points x_i,jCentered 3 x 3 block v_i,jStep 4, obtaining the neighborhood pixel point x_lL =1,2, …,8 flipped 3 × 3 block

5.2) calculating the two pixel blocks v_i,jAndratio r of_i,k：

r_{i, k} = \min {\frac{v_{i, k}}{v_{l, k}}, \frac{v_{l, k}}{v_{i, k}}},

r_i,k∈[0,1]，l=1,2,…,8，k=1,2,…,9，

5.3) calculating pixel point x by using ratio distribution probability formula_i,jRatio r of_i,kProbability of occurrence p (r)_i,k) The calculation is divided into two cases:

if the input SAR image v is a magnitude image, then:

if the input SAR image v is an intensity image:

5.4) defining the pixel point x to be processed_i,jAnd 8 neighborhood pixel points x thereof_lThe distance d of similarity therebetween_lComprises the following steps:

k=1,2,…,9，l=1,2,…,8

respectively calculate pixel points x_i,jAnd 8 neighborhood pixel points x thereof_lThe similarity distance d of_l。

Step 6, the similarity distance d obtained in the step 5₁,d₂,…,d₈Sorted in ascending order, the sorted result isGet the distance after the sequence asAs pixel point x_i,jLike the pixel point x_i,jAnd the mean value of the gray values of the 3 similar points is taken as a pixel point x_i,jThe preprocessed gray value.

And 7, repeating the step 2 to the step 6, and calculating the recovery values of all the pixel points to obtain the preprocessed image.

The effect of the invention can be further confirmed by the following experiments:

experimental conditions and contents

The experimental conditions are as follows: the input image used for the experiment was a two-view X-band amplitude SAR image (Bedfordshire), as shown in fig. 2.

The experimental method comprises the following steps: under the above experimental conditions, two typical algorithms for image preprocessing and the method of the present invention are selected for comparison experiments, which are respectively: (1) minimum Mean Square Error (MMSE) method; (2) and (4) a mean value method.

Second, the experimental contents

Experiment one: fig. 2 is processed by the existing MMSE method, wherein the block size is 3 × 3, and the experimental results are shown in fig. 3 and fig. 6, wherein fig. 6 is a detail enlarged view of fig. 3. As can be seen from fig. 3 and fig. 6, the MMSE method has good noise suppression capability in a homogeneous region, but the edges and the details are blurred, and the brightness of the strong reflection point target is compressed.

Experiment two: fig. 2 is processed by the prior art averaging method, wherein the block size is 3 × 3, and the experimental results are shown in fig. 4 and fig. 7, wherein fig. 7 is a detail enlarged view of fig. 4. As can be seen from the result graphs of fig. 7 and fig. 4, the noise suppression capability stability of the mean method is better than that of the MMSE method, but the edge and texture information of the image cannot be well maintained, and the brightness of the target with strong reflection points is severely compressed.

Experiment three: the method of the invention is used for processing the image of the figure 2, wherein the block size is 3 multiplied by 3, the experimental result is shown in figures 5 and 8, wherein figure 8 is a detail enlarged view of figure 5, and the result graphs of figures 8 and 5 show that the method of the invention can effectively inhibit image noise, well maintain the edge and texture information of the image and well maintain the strong reflection point target.

The experimental results show that the method is superior to other two similar preprocessing methods in overall performance, can well smooth noise and simultaneously keep the details of edges, textures and the like of a natural image, and has good retentivity on a strong reflection point target.

Claims

1. A SAR image preprocessing method based on similar blocks comprises the following steps:

(3) will be provided withWith the pixel point x_i,jTaking the average value of the pixels in the 3 x 3 blocks as the center as the pixel value of the pixel point after preprocessing;

(4) for pixel point x_i,j8 neighborhood pixel point x_l1,2, …,8, and takes pixel x_lCentered 3 x 3 block v_lTo block v_lBy rotating to make pixel point x_i,jCentered 3 x 3 block v_i,jAnd with pixel point x_lCentered 3 x 3 block v_lThe medium homogeneous region is at the same position, v_lThe rotated block is marked as

(5) Calculating a pixel point x_i,jAnd 8 neighborhood pixel points x thereof_l1,2, …,8 similarity distance based on block ratio probability:

5a) taking pixel x_i,jCentered 3 x 3 block v_i,jObtaining the neighborhood pixel point x in the step (4)_l1,2, …,8 inverted 3 × 3 blocks

5b) Calculating the two pixel blocks v_i,jAndratio r of_i,k：

(6) for the similarity distance d obtained in the step (5)₁,d₂,…,d₈Sorted in ascending order, the sorted result isGet the distance after the sequence asAs pixel point x_i,jLike the pixel point x_i,jAnd the mean value of the gray values of the 3 similar points is taken as a pixel point x_i,jPreprocessing the gray value;

(7) and (5) repeating the steps (2) to (6), and calculating the recovery values of all the pixel points to obtain the preprocessed image.

2. The SAR image preprocessing method based on similar blocks as claimed in claim 1, wherein said calculating variance coefficient CV of all pixel points in step (1) to obtain variance coefficient matrix K is performed according to the following steps:

1a) pixel point x_i,j，i∈[1,m],j∈[1,n]Coefficient of variance CV of_i,jThe calculation formula of (2) is as follows:

1b) calculating the variance coefficient of each pixel point in the SAR image v to obtain a variance coefficient matrix K₀：

K₀＝{CV_i,j},i∈[1,m],j∈[1,n]；

1c) For the calculated variance coefficient matrix K₀And carrying out 3 × 3 mean filtering to obtain a filtered variance coefficient matrix K.

3. The SAR image preprocessing method based on similar blocks as claimed in claim 1, wherein said setting variance coefficient classification threshold T in step (2)_cvThreshold value T_cvCalculated according to the following formula:

2a) if the input is an L-view amplitude SAR image, the threshold value T_cvCalculated by the following formula:

2b) if the input is the L apparent intensity SAR image, the threshold value T_cvCalculated by the following formula:

4. the SAR image preprocessing method based on similar blocks as claimed in claim 1, wherein said calculating pixel point x using ratio distribution probability formula in step (5c)_iRatio r of_i,kProbability of occurrence p (r)_i,k) The calculation is performed in two cases:

if the input SAR image v is a magnitude image, its probability p (r)_i,k) Calculated by the following ratio distribution probability formula:

if the input SAR image v is an intensity image, its probability p (r)_i,k) Calculated by the following ratio distribution probability formula: