CN103455975B - The method that diameter radar image edge strengthens - Google Patents

Abstract

The invention discloses a kind of method that diameter radar image edge strengthens, comprising: its normalization average value ratio is calculated for each pixel in synthetic-aperture radar original image; In synthetic-aperture radar original image, the normalization average value of each pixel is than the gray-scale value as this pixel, repaints diameter radar image.In method of the present invention, applying serial mean than optimizing the gray-scale value obtaining respective element, redrawing diameter radar image, enhancing the edge strength of diameter radar image.

Description

The method that diameter radar image edge strengthens

Technical field

The present invention relates to radar image processing technology field, particularly relate to a kind of method that diameter radar image edge strengthens.

Background technology

Along with the application of synthetic-aperture radar (SyntheticApertureRadar is called for short SAR) image is more and more extensive, the marginal information how extracting diameter radar image becomes the key issue of diameter radar image interpretation decipher.Strengthen visual effect by strengthening diameter radar image edge, be beneficial to subsequent applications, this has become an important content of diameter radar image process.

Diameter radar image is owing to being coherent imaging, the speckle noise making it intrinsic have impact on the marginal information on image greatly, therefore conventional Edge extraction algorithm is difficult to carry out effective edge extraction, as the direct spatial domain method processed original image, comprise laplacian spectral radius, Image multiscale edges enhancing etc., and certain conversion is carried out to original image, as wavelet transformation etc., carry out processing to realize image enhaucament at transform domain.Laplacian spectral radius as image utilizes Laplace operator to carry out a kind of method of edge enhancing to image, Laplace operator is in Image neighborhood based on pixel grey scale Difference Calculation, and a kind of Image neighborhood derived by second-order differential strengthens algorithm.Its basic thought is, when average gray lower than other pixels in its place neighborhood of the center pixel gray scale of neighborhood, the gray scale of this center pixel should be further reduced, when average gray higher than other pixels in its place neighborhood of the center pixel gray scale of neighborhood, the gray scale of this center pixel should be further improved, and realizes the Edge contrast of image with this.In algorithm realization process, laplacian spectral radius algorithm is by four directions to centre of neighbourhood pixel or from all directions to asking gradient, and gradient and phase Calais are judged the relation of other pixel grey scales in center pixel gray scale and neighborhood, and by the result of gradient algorithm, pixel grey scale is adjusted.

Visible, said method of the prior art is all the amplitude promoting original high fdrequency component in image, do not produce new radio-frequency component, and the speckle noise of diameter radar image is typical multiplicative characteristic, these edge enhancing method effects are very limited, cannot obtain satisfied edge and strengthen result.

Summary of the invention

(1) technical matters that will solve

For solving above-mentioned one or more problems, the invention provides a kind of method that diameter radar image edge strengthens, to improve the effect that edge strengthens.

(2) technical scheme

According to an aspect of the present invention, provide a kind of method that diameter radar image edge strengthens, comprising: for each pixel in synthetic-aperture radar original image, calculate its normalization average value ratio; In synthetic-aperture radar original image, the normalization average value of each pixel is than the gray-scale value as this pixel, repaints diameter radar image.

(3) beneficial effect

As can be seen from technique scheme, diameter radar image edge enhancing method of the present invention has following beneficial effect: the otherness between the ROA that the present invention utilizes different windows size to calculate is to strengthen the marginal information of diameter radar image, the further sharpening edge strength of diameter radar image, improve the effect that edge strengthens, be conducive to rim detection and the extraction of diameter radar image.

Accompanying drawing explanation

Fig. 1 is the process flow diagram of embodiment of the present invention diameter radar image edge enhancing method;

Fig. 2 is the schematic diagram building window in the average ratio step calculating diameter radar image in embodiment of the present invention diameter radar image edge enhancing method; Wherein, Fig. 2 A is the schematic diagram of horizontal direction, and Fig. 2 B is the schematic diagram of vertical direction; Fig. 2 C is the schematic diagram in left-leaning direction; Fig. 2 D is the schematic diagram in Right deviation direction.

Fig. 3 is the original diameter radar image inputted in embodiment of the present invention diameter radar image edge enhancing method;

Fig. 4 calculates the average ratio that window size is 3 in embodiment of the present invention diameter radar image edge enhancing method;

Fig. 5 calculates the average ratio that window size is 5 in embodiment of the present invention diameter radar image edge enhancing method;

Fig. 6 calculates the average ratio that window size is 7 in embodiment of the present invention diameter radar image edge enhancing method;

Fig. 7 be use the present invention execute the original diameter radar image of routine diameter radar image edge enhancing method to Fig. 3 process after diameter radar image.

Embodiment

For making the object, technical solutions and advantages of the present invention clearly understand, below in conjunction with specific embodiment, and with reference to accompanying drawing, the present invention is described in more detail.

It should be noted that, in accompanying drawing or instructions describe, similar or identical part all uses identical figure number.In the accompanying drawings to simplify or convenient sign, and the implementation not illustrating in accompanying drawing or describe, be form known to a person of ordinary skill in the art in art.In addition, although herein can providing package containing the demonstration of the parameter of particular value, should be appreciated that, parameter without the need to definitely equaling corresponding value, but can be similar to corresponding value in acceptable error margin or design constraint.

The method that diameter radar image edge of the present invention strengthens, different windows size is utilized to calculate the average ratio (RatioOfAverages of the correspondence of diameter radar image, be called for short ROA), then the ROA of these different windows is combined, calculate the best normalization differential index (di) on same picture position, extract best normalization differential index (di) and carry out sharpening, thus the edge realizing diameter radar image strengthens.

In one exemplary embodiment of the present invention, propose a kind of process flow diagram of diameter radar image edge enhancing method.Fig. 1 is the process flow diagram of embodiment of the present invention diameter radar image edge enhancing method.As shown in Figure 1, the present embodiment comprises the following steps:

Step S102, for each pixel in synthetic-aperture radar original image, calculates its normalization average value ratio;

Wherein, for the pixel of in synthetic-aperture radar original image, calculate its normalized step and comprise further:

Step S102a, utilize the square window of N number of different size to calculate the average ratio of this pixel, obtain the average ratio sequence based on different size square window, wherein N is for being greater than 2 integers;

Wherein, for a pixel in synthetic-aperture radar original image, with square window size for w1 calculates its average ratio R _w1:

$R_{w1}=\min \left(R_{w1}^k\right)---\left(1\right)$

$\begin{array}{cc}{R}_{w1}^k=\min (P_{w1}^k/Q_{w1}^k,Q_{w1}^k/P_{w1}^k)& k=\mathrm{1,2,3,4}---\left(2\right)\end{array}$

In formula (1) and (2), represent point average ratio that two side areas dividing mode calculates in kth, min () expression gets wherein minimum one, be the average gray value of this pixel two side areas pixel respectively, k represents the dividing mode of two side areas.

Fig. 2 calculates the schematic diagram building window in average ratio step in embodiment of the present invention diameter radar image edge enhancing method.As shown in Figure 2, centered by this pixel, size is that the square window of w1 × w1 pixel is divided into two adjacent and zero lap regions by it.Wherein:

During k=1, $\begin{array}{c}{R}_{w1}^1=\min (P_{w1}^1/Q_{w1}^1,Q_{w1}^1/P_{w1}^1),\end{array}$ upper-side area pixel average gray value and the underside area pixel average gray value of this pixel respectively, as shown in Figure 2 A;

During k=2, $\begin{array}{c}{R}_{w1}^2=\min (P_{w1}^2/Q_{w1}^2,Q_{w1}^2/P_{w1}^2),\end{array}$ left field pixel average gray value and the right side area pixel average gray value of this pixel respectively, as shown in Figure 2 B;

During k=3, $R_{w1}^3=\min (P_{w1}^3/Q_{w1}^3,Q_{w1}^3/P_{w1}^3),$ upper right side area pixel average gray value and the lower left side area pixel average gray value of this pixel respectively, as shown in Figure 2 C;

During k=4, $\begin{array}{c}{R}_{w1}^k=\min (P_{w1}^4/Q_{w1}^4,Q_{w1}^4/P_{w1}^4),\end{array}$ upper left side area pixel average gray value and the lower right side area pixel average gray value of this pixel respectively, as shown in Figure 2 D.

Except the two side areas dividing mode that Fig. 2 provides, those skilled in the art it will also be appreciated that and other Region dividing mode will not enumerate herein.Further, Region dividing mode herein also can be three kinds, five kinds etc., also no longer describe in detail herein.

Fig. 3 is the original diameter radar image of input.Fig. 4-Fig. 6 is the different windows size ROA value after using the original diameter radar image of the embodiment of the present invention to Fig. 3 to process, Fig. 4 to be calculation window size be 3 ROA; Fig. 5 to be calculation window size be 5 ROA; Fig. 6 to be calculation window size be 7 ROA.

In this step, get w1 and be different size and be greater than the odd number numerical value of 1, and obtain the window of different size, just can calculate a series of ROA value to each pixel.In this step, w1 at least needs to get two values, but preferably should get 3 values, and the span of w1 is 3,5,7,9,11,13,15 etc., and preferably, N=3, w1 value is 3,5,7.The average ratio sequence obtained can be expressed as: [R ₃, R ₅, R ₇].

Step S102b, chooses the maximal value R of average ratio in described average ratio sequence _maxwith minimum value R _min;

Step S102c, utilizes the maximal value R of this average ratio _maxwith minimum value R _min, calculate the best normalization differential index (di) of this pixel:

$I(x,y)=\frac{R_{\max }-R_{\min }}{R_{\max }+R_{\min }}---\left(3\right)$

Wherein, x represents the lateral attitude of pixel in image, and y represents the lengthwise position of pixel in image.

Step S102d: according to described best normalization differential index (di), calculate to should pixel normalization average value ratio:

E(x，y)＝[1.0-I(x，y)]×255.0(4)

Step S104, in synthetic-aperture radar original image, the normalization average value of each pixel is than the gray-scale value as this pixel, repaints diameter radar image.

Fig. 7 is the diameter radar image after using the original diameter radar image of embodiment of the present invention diameter radar image edge enhancing method to Fig. 3 to process.As shown in Figure 7, the diameter radar image after process, the acutance at its edge has been got back than Fig. 4-Fig. 6 and has been improved greatly.

Method of the present invention contributes to promoting diameter radar image rim detection and extraction, can be widely used in the fields such as diameter radar image coupling, atural object contours extract and identification.

Above-described specific embodiment; object of the present invention, technical scheme and beneficial effect are further described; be understood that; the foregoing is only specific embodiments of the invention; be not limited to the present invention; within the spirit and principles in the present invention all, any amendment made, equivalent replacement, improvement etc., all should be included within protection scope of the present invention.

Claims (6)

1. a method for diameter radar image edge enhancing, comprising:

For each pixel in synthetic-aperture radar original image, adopt following methods, calculate its normalization average value ratio:

N number of different edge little square window of growing up is utilized to calculate the average ratio of this pixel, the average ratio sequence of the little square window that obtains growing up based on different edge;

The maxima and minima of average ratio is chosen in described average ratio sequence;

Utilize the maxima and minima of described average ratio, calculate the best normalization differential index (di) of this pixel; And

According to described best normalization differential index (di), calculate the normalization average value ratio of this pixel; And

In synthetic-aperture radar original image, the normalization average value of each pixel is than the gray-scale value as this pixel, repaints diameter radar image.

2. the method for diameter radar image edge according to claim 1 enhancing, is the square window of w1 pixel for length of side size, according to following formula, calculates the average ratio R of this pixel _w1:

$R_{w1}=\min \left(R_{w1}^k\right)$

$R_{w1}^k=min(P_{w1}^k/Q_{w1}^k,Q_{w1}^k/P_{w1}^k)$

Wherein, min () expression gets wherein minimum one, this pixel two side areas pixel average gray value under kth kind two side areas dividing mode respectively, represent point average ratio calculated in kth kind two side areas dividing mode, k represents the dividing mode of this pixel two side areas.

3. the method for diameter radar image edge according to claim 2 enhancing, described k=1,2,3,4;

$R_{w1}=min(R_{w1}^1,R_{w1}^2,R_{w1}^3,R_{w1}^4)$

Wherein:

During k=1, $R_{w1}^1=min(P_{w1}^1/Q_{w1}^1,Q_{w1}^1/P_{w1}^1),$ upper-side area pixel average gray value and the underside area pixel average gray value of this pixel respectively;

During k=2, $R_{w1}^2=min(P_{w1}^2/Q_{w1}^2,Q_{w1}^2/P_{w1}^2),$ left field pixel average gray value and the right side area pixel average gray value of this pixel respectively;

During k=3, $R_{w1}^3=\min (P_{w1}^3/Q_{w1}^3,Q_{w1}^3/P_{w1}^3),$ upper right side area pixel average gray value and the lower left side area pixel average gray value of this pixel respectively;

During k=4, $R_{w1}^k=min(P_{w1}^4/Q_{w1}^4,Q_{w1}^4/P_{w1}^4),$ upper left side area pixel average gray value and the lower right side area pixel average gray value of this pixel respectively.

4. the method for diameter radar image edge according to claim 2 enhancing, wherein, described N=3, described w1=3,5,7.

5. the method for diameter radar image edge according to any one of claim 1 to 4 enhancing, adopts following formula, calculates the best normalization differential index (di) of this pixel:

$I(x,y)=\frac{R_{max}-R_{\min }}{R_{max}+R_{\min }}$

Wherein, R _maxfor the maximal value of average ratio, R _minfor the minimum value of average ratio.

6. the method for diameter radar image edge according to claim 5 enhancing, adopts following formula, calculates the normalization average value ratio of this pixel:

E(x,y)＝[1.0-I(x,y)]×255.0。