A target damage effect evaluation method based on hyperspectral polarization imaging

Abstract

The application discloses a target damage effect evaluation method based on hyperspectral polarization imaging, and comprises the following contents: collecting hyperspectral polarization images of a target before and after damage in multiple wave bands; pre-processing spectral polarization images of the target before and after damage, including high-precision registration of the images before and after damage, polarization image restoration in multiple wave bands, and fusion of a degree of polarization (DOP) image; extracting texture features of the DOP images of the target before and after damage, and calculating texture feature changes of the target before and after damage; and finally, comprehensively evaluating the damage effect of the target according to the texture feature changes before and after damage. The application can integrate spectral, polarization and image information of the target, and accurately evaluate the damage effect of the target through a gray level co-occurrence matrix.

Description

Technical Field

[0001] This invention belongs to the field of spectral polarization detection technology, and in particular, it is a method for evaluating target damage effects based on hyperspectral polarization imaging. Background Technology

[0002] Target damage effect assessment is an important basis for the development of modern weapon systems. Accurate target damage effect assessment is an essential part of formulating response plans and configuring attack weapons in on-site operations, and it is also an important basis for upgrading and optimizing weapon technology.

[0003] Methods for assessing target damage effects can be broadly categorized into three types: manual interpretation methods, which are simple and reliable but require human intervention and are time-consuming; damage assessment methods based on Bayesian theory, which can fully utilize various types of information but rarely rely on human interpretation knowledge and experience; and damage assessment methods based on image analysis, which fully utilize human experience as prior knowledge and have fast image processing algorithms, but due to the complex and ever-changing battlefield environment, traditional detection techniques based on target light intensity information are difficult to meet the requirements, leading to unstable assessment results. Hyperspectral polarization imaging technology, as an advanced optical measurement method, integrates the detection techniques of spectral information, polarization information, and two-dimensional image information of the target medium, providing a richer information source for target identification and analysis, and thus improving the ability to assess target damage effects. Summary of the Invention

[0004] The purpose of this invention is to address the problem that traditional detection techniques based on target light intensity information carry limited information by providing a hyperspectral polarization imaging detection method for evaluating the damage effect on a target.

[0005] The technical solution to achieve the objective of this invention is as follows: On the one hand, a method for evaluating target damage effect based on hyperspectral polarization imaging is provided, the method comprising the following steps:

[0006] Step 1: Acquire hyperspectral polarization images of the target before and after damage;

[0007] Step 2: Preprocess the hyperspectral polarization images of the target before and after damage;

[0008] Step 3: Extract the texture features of the preprocessed image using the gray-level co-occurrence matrix method, and calculate the changes in texture features before and after the target damage;

[0009] Step 4: Evaluate the damage effect on the target by comprehensively considering the changes in texture features before and after the target is damaged.

[0010] Furthermore, the acquisition of hyperspectral polarization images of the target before and after damage in step 1 is specifically achieved using a hyperspectral polarization imaging system based on liquid crystal modulation.

[0011] Furthermore, the preprocessing of the hyperspectral polarization images before and after target damage described in step 2 includes the following specific steps:

[0012] Step 2-1: Use the local upsampling phase correlation method to perform image registration on the hyperspectral polarization images of the target before and after damage;

[0013] Step 2-2: For the registered hyperspectral polarization image, use the full polarization restoration algorithm to restore the Stokes vector images S0, S1, S2, and S3 for each band, and then calculate the DOP image of the target.

[0014]

[0015] Steps 2-3: Fuse the DOP images of the target before and after damage across all bands:

[0016]

[0017] In the formula, M and N represent the width and height of the image, respectively, n represents the number of selected spectral channels, and DOP... k (i,j) represents the value of (i,j) of each pixel in the DOP image under the k-th spectral channel.

[0018] Furthermore, step 3 involves extracting the texture features of the preprocessed image using the gray-level co-occurrence matrix method and calculating the texture changes before and after target damage. The specific process includes:

[0019] Step 3-1: Compress the grayscale level of the DOP image to 32;

[0020] Step 3-2: Determine the size and step size of the sliding window. The sliding window traverses the DOP image in four directions: 0°, 45°, 90°, and 135°, and calculates the gray-level co-occurrence matrix in the four directions.

[0021] Step 3-3: Calculate the gray-level co-occurrence matrix texture features, including contrast, correlation, energy, and inverse difference. Take the average of the texture features calculated in the four directions of 0°, 45°, 90°, and 135° as the final texture features.

[0022] Steps 3-4: Calculate the weight of each texture feature using the following formula:

[0023]

[0024] In the formula, V a W represents the coefficient of variation of the a-th texture feature.a This represents the weight of the a-th texture feature;

[0025]

[0026] In the formula, σ a ,u a Let represent the standard deviation and mean of the a-th texture feature, respectively;

[0027] Steps 3-5: Calculate the Euclidean distance of each pixel of each texture feature before and after the target is damaged. The specific formula is as follows:

[0028]

[0029] In the formula, d a (i,j) represents the Euclidean distance of each pixel (i,j) of the a-th texture feature, T 1a (i,j) and T 2a (i,j) represent the values ​​of each pixel (i,j) of the a-th texture feature before and after the target is damaged;

[0030] Steps 3-6: Calculate the total texture feature change, using the following formula:

[0031]

[0032] In the formula, m and n represent the width and height of the DOP image.

[0033] Furthermore, step 4, which involves evaluating the damage effect on the target by considering the changes in texture features before and after the target is damaged, specifically includes:

[0034] The degree of damage to the target is defined into five levels, as shown below:

[0035] Undamaged, D∈[0,0.05];

[0036] Minor damage, D∈[0.05,0.15];

[0037] Moderate damage, D∈[0.15,0.25];

[0038] Severe damage, D∈[0.25,0.4];

[0039] Scrapped, D≥0.4.

[0040] Compared with the prior art, the significant advantages of this invention are:

[0041] 1) Hyperspectral polarization images are multidimensional information bodies that integrate the spectral information, polarization information, and image information of the target medium, which is beneficial to improving the assessment capability of target damage effects.

[0042] 2) The overall solution can avoid interference from camouflaged targets, making target detection and identification more effective.

[0043] The present invention will now be described in further detail with reference to the accompanying drawings. Attached Figure Description

[0044] Figure 1 This is a flowchart of the target damage effect evaluation method based on hyperspectral polarization imaging according to the present invention.

[0045] Figure 2 This is a schematic diagram of the structure of a hyperspectral polarization imaging device in one embodiment.

[0046] Figure 3 This is a flowchart of spectral polarization image preprocessing in one embodiment.

[0047] Figure 4 This is a flowchart of texture feature extraction using the Gray-Level Co-occurrence Matrix (GLCM) method in one embodiment. Detailed Implementation

[0048] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0049] It should be noted that if the embodiments of the present invention involve directional indicators (such as up, down, left, right, front, back, etc.), the directional indicators are only used to explain the relative positional relationship and movement of the components in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indicators will also change accordingly.

[0050] Furthermore, if the embodiments of this invention involve descriptions such as "first" or "second," these descriptions are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined with "first" or "second" may explicitly or implicitly include at least one of those features. Additionally, the technical solutions of the various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. If the combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed by this invention.

[0051] In one embodiment, combined Figure 1 This paper provides a method for evaluating target damage effects based on hyperspectral polarization imaging, the method comprising the following steps:

[0052] Step 1: Acquire hyperspectral polarization images of the target before and after damage;

[0053] Step 2: Preprocess the hyperspectral polarization images of the target before and after damage;

[0054] Step 3: Extract the texture features of the preprocessed polarization degree (DOP) image using the gray-level co-occurrence matrix method (GLCM), and calculate the changes in texture features before and after the target damage;

[0055] Step 4: Evaluate the damage effect on the target by comprehensively considering the changes in texture features before and after the target is damaged.

[0056] Furthermore, in one embodiment, the acquisition of hyperspectral polarization images of the target before and after damage in step 1 is specifically implemented using a liquid crystal modulation-based hyperspectral polarization imaging system.

[0057] Preferably, combined with Figure 2 The liquid crystal modulation-based hyperspectral polarization imaging system includes a front imaging objective 1, an aperture 2, a collimating objective 3, a polarization modulation module 4, a spectral tuning module 5, a rear imaging objective 6, and an area array detector 7 arranged sequentially along the optical axis. After the incident light passes through the front imaging objective 1, the aperture 2, the collimating objective 3, and the polarization modulation module 4, four different polarization states of light can be generated within one modulation cycle. After passing through the spectral tuning module 5 and the rear imaging objective 6, the light is finally imaged on the area array detector 7, and spectral polarization images with four different polarization states under different wavelengths are acquired.

[0058] The polarization modulation module 4 includes a ferroelectric liquid crystal 4.1, a half-wave plate 4.2, a ferroelectric liquid crystal 4.3, and a quarter-wave plate 4.4 arranged sequentially along the optical axis.

[0059] Preferably, the spectral tuning module 5 is a liquid crystal tunable filter.

[0060] Furthermore, in one embodiment, the set Figure 3 Step 2 involves preprocessing the hyperspectral polarization images of the target before and after damage. The specific process includes:

[0061] Step 2-1: Use the local upsampling phase correlation method to perform image registration on the hyperspectral polarization images of the target before and after damage;

[0062] Step 2-2: For the registered hyperspectral polarization image, use the full polarization restoration algorithm to restore the Stokes vector images S0, S1, S2, and S3 for each band, and then calculate the DOP image of the target.

[0063]

[0064] Steps 2-3: Fuse the DOP images of the target before and after damage across all bands:

[0065]

[0066] In the formula, M and N represent the width and height of the image, respectively, n represents the number of selected spectral channels, and DOP... k (i,j) represents the value of (i,j) of each pixel in the DOP image under the k-th spectral channel.

[0067] Furthermore, in one embodiment, combined with Figure 4 Step 3 involves extracting the texture features of the preprocessed image using the gray-level co-occurrence matrix method and calculating the texture changes before and after the target damage. The specific process includes:

[0068] Step 3-1: Compress the grayscale level of the DOP image to 32;

[0069] Step 3-2: Select a sliding window of size 7×7 with a step size of 1. Slide the window to traverse the DOP image in four directions: 0°, 45°, 90°, and 135° and calculate the gray-level co-occurrence matrix in the four directions.

[0070] Step 3-3: Calculate the gray-level co-occurrence matrix texture features, including contrast, correlation, energy, and inverse difference. Take the average of the texture features calculated in the four directions of 0°, 45°, 90°, and 135° as the final texture features.

[0071] Steps 3-4: Calculate the weight of each texture feature using the following formula:

[0072]

[0073] In the formula, V a W represents the coefficient of variation of the a-th texture feature. a This represents the weight of the a-th texture feature;

[0074]

[0075] In the formula, σ a ,u a Let represent the standard deviation and mean of the a-th texture feature, respectively;

[0076] Steps 3-5: Calculate the Euclidean distance of each pixel of each texture feature before and after the target is damaged. The specific formula is as follows:

[0077]

[0078] In the formula, d a (i,j) represents the Euclidean distance of each pixel (i,j) of the a-th texture feature, T 1a (i,j) and T 2a(i,j) represent the values ​​of each pixel (i,j) of the a-th texture feature before and after the target is damaged;

[0079] Steps 3-6: Calculate the total texture feature change, using the following formula:

[0080]

[0081] In the formula, m and n represent the width and height of the DOP image.

[0082] Furthermore, in one embodiment, step 4, which involves evaluating the damage effect on the target by synthesizing the changes in texture features before and after the target damage, specifically includes:

[0083] The degree of damage to the target is defined into five levels, as shown below:

[0084] Undamaged, D∈[0,0.05];

[0085] Minor damage, D∈[0.05,0.15];

[0086] Moderate damage, D∈[0.15,0.25];

[0087] Severe damage, D∈[0.25,0.4];

[0088] Scrapped, D≥0.4.

[0089] In one embodiment, a target damage assessment system based on hyperspectral polarization imaging is provided, the system comprising:

[0090] The first module is used to acquire hyperspectral polarization images of the target before and after damage.

[0091] The second module is used to preprocess the hyperspectral polarization images of the target before and after damage;

[0092] The third module is used to extract the texture features of the preprocessed image using the gray-level co-occurrence matrix method and to calculate the changes in texture features before and after the target is damaged.

[0093] The fourth module is used to comprehensively evaluate the damage effect on the target by considering the changes in texture features before and after the target is damaged.

[0094] Specific limitations regarding the target damage assessment system based on hyperspectral polarization imaging can be found in the limitations of the target damage assessment method based on hyperspectral polarization imaging mentioned above, and will not be repeated here. Each module in the aforementioned target damage assessment system based on hyperspectral polarization imaging can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0095] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to perform the following steps:

[0096] Step 1: Acquire hyperspectral polarization images of the target before and after damage;

[0097] Step 2: Preprocess the hyperspectral polarization images of the target before and after damage;

[0098] Step 3: Extract the texture features of the preprocessed image using the gray-level co-occurrence matrix method, and calculate the changes in texture features before and after the target damage;

[0099] Step 4: Evaluate the damage effect on the target by comprehensively considering the changes in texture features before and after the target is damaged.

[0100] For specific limitations on each step, please refer to the limitations on the target damage effect assessment method based on hyperspectral polarization imaging mentioned above, which will not be repeated here.

[0101] This invention integrates the spectral, polarization, and image information of a target, and uses a gray-level co-occurrence matrix to accurately assess the damage effect on the target.

[0102] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention without departing from its spirit and scope should be included within the protection scope of the present invention.

Claims

1. A method for evaluating target damage effects based on hyperspectral polarization imaging, characterized in that, The method includes the following steps: Step 1: Acquire hyperspectral polarization images of the target before and after damage; Step 2: Preprocess the hyperspectral polarization images of the target before and after damage; Step 3: Extract the texture features of the preprocessed image using the gray-level co-occurrence matrix method, and calculate the changes in texture features before and after the target damage; Step 4: Evaluate the damage effect on the target by comprehensively considering the changes in texture features before and after the target is damaged; Step 2 involves preprocessing the hyperspectral polarization images of the target before and after damage. The specific process includes: Step 2-1: Use the local upsampling phase correlation method to perform image registration on the hyperspectral polarization images of the target before and after damage; Step 2-2: For the registered hyperspectral polarization image, use the full polarization restoration algorithm to restore the Stokes vector images S0, S1, S2, and S3 for each band, and then calculate the DOP image of the target. Steps 2-3: Fuse the DOP images of the target before and after damage across all bands: In the formula, M and N represent the width and height of the image, respectively, n represents the number of selected spectral channels, and DOP... k (i,j) represents the value of (i,j) of each pixel in the DOP image under the k-th spectral channel; Step 3 involves extracting texture features from the preprocessed image using the gray-level co-occurrence matrix method and calculating the texture changes before and after target damage. The specific process includes: Step 3-1: Compress the grayscale level of the DOP image to 32; Step 3-2: Determine the size and step size of the sliding window. The sliding window traverses the DOP image in four directions: 0°, 45°, 90°, and 135°, and calculates the gray-level co-occurrence matrix in the four directions. Step 3-3: Calculate the gray-level co-occurrence matrix texture features, including contrast, correlation, energy, and inverse difference. Take the average of the texture features calculated in the four directions of 0°, 45°, 90°, and 135° as the final texture features. Steps 3-4: Calculate the weight of each texture feature using the following formula: In the formula, V a W represents the coefficient of variation of the a-th texture feature. a This represents the weight of the a-th texture feature; In the formula, σ a ,u a Let represent the standard deviation and mean of the a-th texture feature, respectively; Steps 3-5: Calculate the Euclidean distance of each point of each texture feature before and after the target is damaged. The specific formula is as follows: In the formula, d a (i,j) represents the Euclidean distance of each pixel (i,j) of the a-th texture feature, T 1a (i,j) and T 2a (i,j) represent the values ​​of each pixel (i,j) of the a-th texture feature before and after the target is damaged; Steps 3-6: Calculate the total texture feature change, using the following formula: In the formula, m and n represent the width and height of the DOP image.

2. The target damage effect evaluation method based on hyperspectral polarization imaging according to claim 1, characterized in that, The acquisition of hyperspectral polarization images of the target before and after damage, as described in step 1, is specifically achieved using a hyperspectral polarization imaging system based on liquid crystal modulation.

3. The target damage effect evaluation method based on hyperspectral polarization imaging according to claim 2, characterized in that, The liquid crystal modulation-based hyperspectral polarization imaging system includes a front imaging objective (1), an aperture (2), a collimating objective (3), a polarization modulation module (4), a spectral tuning module (5), a rear imaging objective (6), and an area array detector (7) arranged sequentially along the optical axis. After the incident light passes through the front imaging objective (1), the aperture (2), the collimating objective (3), and the polarization modulation module (4), four different polarization states of light can be generated within one modulation cycle. After passing through the spectral tuning module (5) and the rear imaging objective (6), the light is finally imaged on the area array detector (7), and spectral polarization images with four different polarization states under different wavelengths are acquired.

4. The target damage effect evaluation method based on hyperspectral polarization imaging according to claim 3, characterized in that, The polarization modulation module (4) includes a ferroelectric liquid crystal (4.1), a half-wave plate (4.2), a ferroelectric liquid crystal (4.3), and a quarter-wave plate (4.4) arranged sequentially along the optical axis.

5. The target damage effect evaluation method based on hyperspectral polarization imaging according to claim 4, characterized in that, The spectral tuning module (5) is a liquid crystal tunable filter.

6. The target damage effect evaluation method based on hyperspectral polarization imaging according to claim 5, characterized in that, Step 4, which involves evaluating the changes in texture features before and after target damage, specifically includes: The degree of damage to the target is defined into five levels, as shown below: Undamaged, D∈[0,0.05]; Minor damage, D∈[0.05,0.15]; Moderate damage, D∈[0.15,0.25]; Severe damage, D∈[0.25,0.4]; Scrapped, D≥0.

4.

7. A target damage effect evaluation system based on hyperspectral polarization imaging, using the method described in any one of claims 1 to 6, characterized in that, The system includes: The first module is used to acquire hyperspectral polarization images of the target before and after damage. The second module is used to preprocess the hyperspectral polarization images of the target before and after damage; The third module is used to extract the texture features of the preprocessed image using the gray-level co-occurrence matrix method and to calculate the changes in texture features before and after the target is damaged. The fourth module is used to comprehensively evaluate the damage effect on the target by considering the changes in texture features before and after the target is damaged.

8. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method of any one of claims 1 to 6.

Images