Polarization image edge enhancement calculation method and system

Abstract

The invention belongs to the technical field of image processing, and particularly relates to a polarization image edge enhancement calculation method, which comprises the following steps of: 1, acquiring polarization intensity images in 0-degree, 45-degree, 90-degree and 135-degree directions by using a focal plane polarization camera, resolving to obtain a light intensity image IS0, a first polarization parameter image IS1, a second polarization parameter image IS2, a polarization degree image IDoLP and a polarization angle image IAoP; and step 2, calculating and obtaining an edge feature image IEF, a polarization edge enhanced image IPES and polarization edge information amount # imgabs0 # by using each polarization parameter image obtained in the step S1. According to the method provided by the invention, the advantage information of the polarization image and the light intensity image can be simultaneously and reasonably combined for complementation, so that the difference comparison between the target and the background is more obvious; the contour and details of the target are highlighted. The finally obtained polarization edge enhanced image has more significant edge and contour information, and can be applied to subsequent image fusion, target detection and recognition and other advanced computer vision tasks.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and in particular to a polarization image edge enhancement calculation method and system. Background Art

[0002] Polarization is one of the important characteristics of light. The asymmetry of the vibration direction of light with respect to the propagation direction is called polarization. It is the most obvious sign that distinguishes transverse waves from longitudinal waves. Only transverse waves have polarization. Polarization imaging has obvious advantages over traditional imaging technology. It can obtain the polarization information of the target object while obtaining the light intensity information of the target object, especially the surface or subtle changes that are not easily observed by the human eye. Therefore, polarization imaging technology is widely used in military and civilian fields such as target detection, camouflage recognition, and remote sensing detection.

[0003] Edge information is one of the most important features of the target object. It is the dividing line between the target and the background. In the image, edge information is also an extremely important indicator, which is used to describe the differences between different areas in the image. It is of great significance for tasks such as target detection, segmentation, and classification. Although traditional edge calculation methods can enhance or highlight edge contours and details to a certain extent, the effect on polarized images is not outstanding, and there is relatively little research on edge information of polarized images. Therefore, we propose a polarized image edge enhancement calculation method and system to solve the above problems. Summary of the invention

[0004] 1. Technical issues to be resolved

[0005] In view of the deficiencies in the prior art, the present invention provides a polarization image edge enhancement calculation method and system, which solves the problems raised in the above background technology.

[0006] (II) Technical solution

[0007] In order to achieve the above-mentioned purpose, the present invention specifically adopts the following technical solutions:

[0008] A polarization image edge enhancement calculation method comprises the following steps:

[0009] Step 1: Use a focal plane polarization camera to obtain polarization intensity images at 0°, 45°, 90°, and 135°, and calculate the intensity image I S0 , the first polarization parameter image I S1 , the second polarization parameter image I S2 , Polarization Image I DoLP And the polarization angle image I AoP ;

[0010] Step 2: Use the polarization parameter images obtained in step 1 to calculate and obtain the edge feature image I EF, Polarization edge enhanced image I PES and polarization edge information volume

[0011] Step 3: Based on the obtained images, design a polarization edge enhancement function F and add a gradient weight function ω g , while removing noise and further highlighting edge information;

[0012] Step 4: Calculate and obtain the final polarization edge enhanced image I from the obtained polarization edge enhanced image I PES and the polarization edge enhancement function F PEA .

[0013] Furthermore, the formula for solving and obtaining each polarization parameter image in step 1 is:

[0014]

[0015] In the formula, ε represents AoP, δ represents the ellipticity angle, χ represents the radiation intensity, and DoLP can be expressed as cos2δ.

[0016] Furthermore, the calculation formulas for obtaining the edge feature image I EF , the polarization edge enhanced image I PES and the polarization edge information volume in step 2 are as follows:

[0017]

[0018] R B (φ, I EF ) = R B (φ, I EF )(1 - I EF )

[0019]

[0020] In the formula, g x g y represent two kernels of the edge detection operator, I EF is the edge feature image, citing and improving the polarization distance model, using four angular direction polarization images as inputs to calculate the linear polarization target sensitivity R(φ, I EF ) of light in different directions, where the value of φ is Using the edge feature image I EF to highlight the edge of the target source image, S p is the activation threshold; then calculate the target area activation value ET 1 , ET 2 , ET 3, Similarly, the linear polarization background sensitivity R of light in different directions can be calculated B (φ, I EF ) to obtain the background region activation value EB 1 、EB 2 and EB 3 , and finally the polarization edge enhancement image I PES is calculated; Among them and respectively represent the edge intensity and direction orientation value at (i, j), ω X represents the importance of each source image to the polarization edge enhancement image. The polarization edge information amount calculates the information amount transferred from the source image to the polarization edge enhancement image. Through the polarization edge information amount to reflect the edge difference between the source image and the polarization edge enhancement image, highlighting the structural complexity and edge complexity of the polarization edge enhancement image.

[0021] Further, the calculation formula for designing the polarization edge enhancement function F in step 3 is as follows:

[0022]

[0023] In the formula, Ω represents the number of pixels in the Ω region, p and q are pixel points in the Ω region, W p represents the normalized weight, I represents the input image, μ r and μ s respectively represent the gray value range and spatial domain Gaussian functions, and respectively represent their standard deviations, ω g represents the gradient weight function, which calculates the weight through the ratio between gradient amplitudes, represents the gradient information of the image, and C is a constant that controls the influence of gradient difference on the weight.

[0024] Further, in step 4, the calculation formula for obtaining the polarization edge enhancement image I PEA is:

[0025] I PEA = F(I PES )

[0026] In the formula, I PES is the polarization edge enhancement image, F is the polarization edge enhancement function, and I PEA is the polarization edge enhancement image.

[0027] A system for implementing the polarization image edge enhancement calculation method according to any one of claims 1 to 5, the system includes:

[0028] Polarization parameter calculation module: It is used to collect polarization images in four directions by a split focal plane polarization camera, and calculate and obtain the intensity image I S0 , the first polarization parameter image I S1 , the second polarization parameter image I S2 , the degree of polarization image I DoLP and the polarization angle image I AoP ;

[0029] Edge feature image, polarization edge enhancement image, polarization edge information quantity calculation module: It is used for each polarization parameter image calculated above, and calculates and obtains the edge feature image I EF , the polarization edge enhancement image I PES and obtains the polarization edge information quantity

[0030] Polarization edge enhancement function module: It is used to design a polarization edge enhancement function F based on the obtained images to further highlight the edge information while removing noise;

[0031] Polarization edge enhancement image acquisition module: It is used to calculate and obtain the polarization edge enhancement image I based on the calculated polarization edge enhancement image I PES , the polarization edge enhancement function F PEA .

[0032] (III) Beneficial effects

[0033] Compared with the prior art, the present invention provides a polarization image edge enhancement calculation method and system, which has the following beneficial effects:

[0034] 1. In the present invention, intensity images in four directions are obtained by a split focal plane polarization camera, and the intensity image I S0 , the first polarization parameter image I S1 , the second polarization parameter image I S2 , the degree of polarization image I DoLP and the polarization angle image I AoP are calculated and obtained; then the edge feature image I EF , the polarization edge enhancement image I PES and the polarization edge information quantity are calculated from each polarization parameter image; then a polarization edge enhancement function F is designed, and a gradient weight function ω g is added to further highlight the edge information while removing noise; finally, the polarization edge enhancement image I PEA is obtained. The present invention can make the edge information of the polarization image more prominent, highlight the contour and details of the polarization image, significantly improve the image quality, and can be applied to subsequent military and civilian fields such as image fusion, image enhancement, target detection, and edge detection.

[0035] 2. The present invention calculates a polarization edge enhancement image based on various polarization parameter images, which can make the features of the edges of the polarization image more obvious and better utilize the polarization information of the image. The polarization edge enhancement image calculated by the present invention based on the polarization image and the polarization edge enhancement function makes the edge information of the image more prominent. Compared with the traditional edge calculation method, it is more suitable for polarization images. The image generated by this calculation method is more in line with the visual perception and understanding of the human eye, and the edge information can be intuitively seen to have a more prominent visual effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 It is a flowchart of the polarization edge enhancement calculation method provided by the present invention;

[0037] Figure 2 It is a schematic diagram of the principle of the polarization edge enhancement calculation method provided by the present invention;

[0038] Figure 3 It is the S0 light intensity image provided by an example of the present invention;

[0039] Figure 4 It is the S1 first polarization parameter image provided by an example of the present invention;

[0040] Figure 5 It is the S2 second polarization parameter image provided by an example of the present invention;

[0041] Figure 6 It is the DoLP degree of linear polarization image provided by an example of the present invention;

[0042] Figure 7 It is the AoP angle of linear polarization image provided by an example of the present invention;

[0043] Figure 8 It is the edge feature image IEF provided by an example of the present invention;

[0044] Figure 9 It is the polarization edge enhancement image IPES provided by an example of the present invention;

[0045] Figure 10 It is the polarization edge enhancement image IPEA provided by an example of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0046] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0047] Embodiment 1

[0048] As shown Figure 1 in the figure, a polarization image edge enhancement calculation method provided in Embodiment 1 of the present invention includes the following steps:

[0049] Step 1: Use a split focal plane polarization camera to obtain polarization intensity images in the 0°, 45°, 90°, and 135° directions, and solve to obtain the light intensity image I S0 , the first polarization parameter image I S1 , the second polarization parameter image I S2 , the degree of polarization image I DoLP , and the polarization angle image I AoP ;

[0050] Among them, the formulas for solving and obtaining each polarization parameter image are:

[0051]

[0052]

[0053] In the formula, ε represents AoP, δ represents the ellipticity angle, χ represents the radiation intensity, and DoLP can be expressed as cos2δ.

[0054] Step 2: Use each polarization parameter image obtained in Step 1 to calculate and obtain the edge feature image I EF , the polarization edge enhancement image I PES , and the polarization edge information volume The calculation formulas for obtaining the edge feature image I EF , the polarization edge enhancement image I PES are:

[0055]

[0056] R B (φ, I EF ) = R B (φ, I EF )(1 - I EF )

[0057]

[0058]

[0059] In the formula, g x g y represent two kernels of the Prewitt edge detection operator. The size of the two kernels is 3×3, and the small area of each kernel is represented as zi (i ∈ 1 to 9), I EFis an edge feature image. By referring to and improving the polarization distance model, using polarization images in the directions of 0°, 45°, 90°, and 135° as inputs, the linear polarization target sensitivity R(φ, I EF ) is calculated, where the value of φ is The edge feature image I EF is used to highlight the edges of the target source image, and S p is the activation threshold. Then, the activation value ET of the target region is calculated using the formula 1 , ET 2 , ET 3 . Similarly, the linear polarization background sensitivity R B (φ, I EF ) can be calculated to obtain the activation values EB 1 , EB 2 , and EB 3 . Finally, the polarization edge enhancement image I PES is calculated.

[0060] The present invention innovatively proposes the above calculation formula, and the technical effect is as follows: the light intensity image I S0 , the first polarization parameter image I S1 , the second polarization parameter image I S2 , the degree of polarization image I DoLP , the polarization angle image I AoP , the edge feature image I EF , and the polarization edge enhancement image I PES are as shown in Figures 3 - 9 . The calculated polarization edge enhancement image can make the edge information of the target more obvious and at the same time contain rich polarization information.

[0061] Calculate and obtain the polarization edge information volume The calculation formula is as follows:

[0062]

[0063] In the formula, where and respectively represent the edge intensity and direction orientation value at (i, j), and ω X represents the importance of each source image to the polarization edge enhancement image.

[0064] The present invention innovatively proposes the above calculation formula, and the technical effect is as follows: by calculating the polarization edge information volume PES of the polarization edge enhancement image I The polarization edge information volume calculates the information volume transferred from the source image to the polarization edge enhancement image. Through the polarization edge information volume To reflect the edge difference between the source image and the polarization edge-enhanced image, highlighting the structural complexity and edge complexity of the polarization edge-enhanced image.

[0065] Step 3: Based on the obtained image, design the polarization edge enhancement function F and add the gradient weight function ω g , while removing noise, further highlighting the edge information; the calculation formula is as follows:

[0066]

[0067] In the formula, Ω represents the number of pixels in the Ω region, p and q are pixel points in the Ω region, W p represents the normalized weight, I represents the input image, μ r and μ s respectively represent the Gaussian functions in the gray value domain and the spatial domain, and respectively represent their standard deviations, ω g represents the gradient weight function, which calculates the weight through the ratio between the gradient amplitudes, represents the gradient information of the image, and C is a constant that controls the influence of the gradient difference on the weight.

[0068] Step 4: Calculate and obtain the final polarization edge-enhanced image I PES from the obtained polarization edge-enhanced image I PEA and the polarization enhancement function F; the calculation formula is as follows:

[0069] I PEA = F(I PES )

[0070] In the formula, I PES is the polarization edge-enhanced image, F is the polarization edge enhancement function, and I PEA is the polarization edge-enhanced image.

[0071] Through the innovative proposal of calculating the polarization edge enhancement function F, the present invention finally obtains the polarization edge-enhanced image I PEA , as Figure 10 shown, the technical effect is that: while better retaining the edge information of the image in the gray value domain, spatial domain, and gradient respectively, the polarization edge enhancement function F removes the noise in the image, achieving the effect of image smoothing, making the edge information of the image more prominent, and contributing to the further analysis and understanding of the image.

[0072] Embodiment 2

[0073] As another implementation manner of the present invention, a polarization image edge enhancement calculation method provided by Embodiment 2 of the present invention includes the following steps:

[0074] Step 1: Use a split focal plane polarization camera to obtain polarization intensity images in the directions of 0°, 45°, 90°, and 135°, and obtain the corresponding polarization intensity images I0°, I45°, I90°, I135°; and perform calculations on the polarization intensity images I0°, I45°, I90°, I135°, and use the following calculation formula to obtain the polarization parameter image S 0 、S 1 、S 2 、DoLP and AoP;

[0075]

[0076] In the formula, ε represents AoP, δ represents the ellipticity angle, χ represents the radiation intensity, and DoLP can be expressed as cos2δ.

[0077] Step 2: Calculate and obtain the edge feature image I EF 、the polarization edge enhancement image I PES and the polarization edge information amount The calculation formula is as follows:

[0078]

[0079] R B (φ, I EF ) = R B (φ, I EF )(1 - I EF )

[0080]

[0081]

[0082] In the formula, g x g y represents the two kernels of the Sobel edge detection operator, I EF is the edge feature image, reference and improve the polarization distance model, use the polarization images in the directions of 0°, 45°, 90°, and 135° as inputs, and calculate the linear polarization target sensitivity R(φ, I EF ) of light in different directions, where the value of φ is Adopt the edge feature image I EF to highlight the edge of the target source image, S p is the activation threshold. Then use the formula to calculate the target area activation values ET 1 、ET 2 、ET 3 , and similarly, the linear polarization background sensitivity R B (φ, IEF ) Obtain the background region activation value EB 1 , EB 2 and EB 3 , and finally calculate the polarization edge enhanced image I PES , where and represent the edge intensity and direction orientation value at (i, j) respectively, and ω X represents the importance of each source image to the polarization edge enhanced image. The polarization edge information volume calculates the information volume transferred from the source image to the polarization edge enhanced image. Through the polarization edge information volume to reflect the edge difference between the source image and the polarization edge enhanced image, highlighting the structural complexity and edge complexity of the polarization edge enhanced image.

[0083] Step 3: Based on the obtained images, design the polarization edge enhancement function F, add the gradient weight function ω g , remove noise while further highlighting edge information, and the calculation formula is as follows:

[0084]

[0085]

[0086] In the formula, Ω represents the number of pixels in the Ω region, p and q are pixel points in the Ω region, W p represents the normalized weight, I represents the input image, μ r and μ s represent the gray value domain and spatial domain Gaussian functions respectively, and represent their standard deviations respectively, ω g represents the gradient weight function ω g , and this function calculates the weight through the ratio between the gradient amplitudes, represents the gradient information of the image, and C is a constant that controls the influence of the gradient difference on the weight.

[0087] Step 4: Calculate and obtain the final polarization edge enhanced image I PES from the obtained polarization edge enhanced image I PEA and the polarization edge enhancement function F, and the calculation formula is as follows:

[0088] I PEA = F(I PES )

[0089] In the formula, I PES is the polarization edge enhanced image, F is the polarization edge enhancement function, I PEAIt is a polarization edge-enhanced image.

[0090] Embodiment 3

[0091] Embodiment 3 of the present invention provides a polarization image edge enhancement calculation system, including:

[0092] A polarization parameter calculation module, which is used to collect polarization images in four directions by a focal plane polarization camera, and calculate and obtain an intensity image I S0 , a first polarization parameter image I S1 , a second polarization parameter image I S2 , a degree of polarization image I DoLP and a polarization angle image I AoP ;

[0093] An edge feature image, a polarization edge enhancement image, and a polarization edge information calculation module, which are used for each polarization parameter image calculated above, calculate and obtain an edge feature image I EF , a polarization edge enhancement image I PES and obtain the polarization edge information amount

[0094] A polarization edge enhancement function module, which is used to design a polarization edge information enhancement function F based on the obtained images, and add a gradient weight function ω g , to further highlight the edge information while removing noise;

[0095] A polarization edge enhancement image obtaining module, which is used to calculate and obtain a polarization edge saliency image I PES based on the calculated polarization edge enhancement image I PEA .

[0096] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not detailed in a certain embodiment, reference may be made to the relevant descriptions of other embodiments.

[0097] As creative auxiliary evidence for the claims of the present invention, it is also reflected in the following important aspects:

[0098] 1. The polarization image edge enhancement calculation method is a new polarization image calculation method. Compared with traditional image edge calculation methods, this method can make the dominant information between different source images complementary, make the edge information of the image more prominent, and at the same time filter out the noise around the edge. The prominent edge information has important application prospects in target detection and recognition, classification and segmentation, biotechnology, medical imaging, remote sensing images, and the military field.

[0099] 2. The traditional image edge calculation method mainly targets intensity images. The polarization image edge enhancement calculation method of the present invention comprehensively utilizes the polarization characteristic information of each polarization parameter image S 1 、S 2 、DoLP and AoP, as well as the intensity information of the S 0 image, and reasonably utilizes the edge information and contour details of different amplitude images.

[0100] 3. The polarization image edge enhancement calculation method of the present invention can be applied in the field of polarization image processing. It has a better effect than the traditional edge image calculation method, can make the edge information and details of the polarization image more prominent, and provides a basis for subsequent high-level computer vision task analysis and processing.

[0101] Finally, it should be noted that the above are only the preferred embodiments of the present invention and are not used to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A polarization image edge enhancement calculation method, characterized in that: The steps include: Step 1: Use a focal plane polarization camera to obtain polarization intensity images at 0°, 45°, 90°, and 135°, and calculate the intensity image I S0 , the first polarization parameter image I S1 , the second polarization parameter image I S2 , Polarization Image I DoLP And the polarization angle image I AoP ; Step 2: Use the polarization parameter images obtained in step 1 to calculate and obtain the edge feature image I EF , polarization edge enhanced image I PES And the amount of polarization edge information Step 3: Based on the obtained image, design the polarization edge enhancement function F and add the gradient weight function ω g , while removing noise, further highlighting edge information; Step 4: Obtain polarized edge-enhanced image I PES And the polarization edge enhancement function F is calculated to obtain the final polarization edge enhanced image I PEA .

2. The method for calculating edge enhancement of a polarization image according to claim 1, characterized in that: The formula for calculating and obtaining each polarization parameter image in step 1 is: Where ε represents AoP, δ represents the ellipticity angle, χ represents the radiation intensity, and DoLP can be expressed as cos2δ.

3. The method for calculating edge enhancement of a polarization image according to claim 2, characterized in that: In step 2, edge feature image I is calculated and obtained. EF , polarization edge enhanced image I PES And the amount of polarization edge information The calculation formula is as follows: R B (φ,I EF )=R B (φ,I EF )(1-I EF ) In the formula, g x g y Represents the two kernels of the edge detection operator, I EF For edge feature images, the polarization distance model is referenced and improved, and the polarization images in four angular directions are used as input to calculate the linear polarization target sensitivity R(φ,I EF ), where the value of φ is Using edge feature image I EF To highlight the edge of the target source image, S p is the activation threshold; Then the target area activation values ​​ET1, ET2, and ET3 can be calculated using the formula: B (φ,I EF ) to obtain the background area activation values ​​EB1, EB2 and EB3, and finally calculate the polarization edge enhanced image I PES ; in and Respectively represent the edge strength and direction orientation value at (i, j), ω X Represents the importance of each source image to the polarization edge enhanced image.

4. The method for calculating edge enhancement of a polarization image according to claim 1, characterized in that: The calculation formula for the polarization edge enhancement function F designed in step 3 is as follows: Where Ω represents the number of pixels in the Ω area, p and q are the pixel points in the Ω area, and W p represents the normalized weight, I represents the input image, μ r and μ s Represent the gray value domain and spatial domain Gaussian function respectively, and Represent their standard deviations, ω g represents the gradient weight function, Represents the gradient information of the image, and C is a constant.

5. The method for calculating edge enhancement of a polarization image according to claim 4, characterized in that: In step 4, a polarization edge enhanced image I is obtained. PEA The calculation formula is: I PEA =F(I PES ) In the formula, I PES is the polarization edge enhanced image, F is the polarization edge enhancement function, I PEA Enhances images for polarized edges.

6. A system for implementing the polarization image edge enhancement calculation method according to any one of claims 1 to 5, characterized in that: The system includes: Polarization parameter calculation module: used to collect polarization images in four directions from the focal plane polarization camera and calculate the intensity image I S0 , the first polarization parameter image I S1 , the second polarization parameter image I S2 , Polarization Image I DoLP And the polarization angle image I AoP ; Edge feature image, polarization edge enhancement image, polarization edge information calculation module: used for calculating the edge feature image I from the polarization parameter images solved above EF , polarization edge enhanced image I PES And obtain the polarization edge information Polarization edge enhancement function module: used to design a polarization edge enhancement function F based on the acquired image, to remove noise and further highlight edge information; Polarization edge enhanced image acquisition module: used to enhance the polarization edge image I based on the calculated PES , polarization edge enhancement function F, calculate and obtain the polarization edge enhanced image I PEA .

Images