Hyperspectral abnormal target detection method and device based on band index

Abstract

The invention provides a hyperspectral abnormal target detection method and device based on a band index, and relates to hyperspectral data processing, and the method comprises the steps: calculating the band index of an original hyperspectral image; arranging the waveband indexes from large to small, selecting the first n wavebands as optimal wavebands, and forming a new data cube by the wavebands to obtain a new hyperspectral image; and performing abnormal target detection on the new hyperspectral image by using an RX algorithm to obtain an abnormal detection result. According to the scheme, the problems of low target detection precision and low calculation speed caused by large redundant information between bands of the hyperspectral image can be effectively solved.

Description

Technical Field

[0001] The present invention relates to the technical field of hyperspectral data processing, and in particular to a hyperspectral abnormal target detection method and device based on band index. Background Art

[0002] The RX algorithm was first proposed by Reed and XiaoliYu. It has achieved remarkable success in detecting abnormal targets in multispectral and hyperspectral images. The algorithm has strong adaptability, simple structure, easy implementation, good detection performance, and is one of the most widely used anomaly detection methods.

[0003] In the related art, since hyperspectral images have extremely high spectral resolution and number of bands, it is necessary to calculate the covariance matrix of a large number of samples when directly using the RX algorithm to process hyperspectral images, which leads to a high false alarm probability in the detection results.

[0004] Based on this, there is an urgent need for a hyperspectral abnormal target detection method and device based on band index to solve the above technical problems. Summary of the invention

[0005] The embodiments of the present invention provide a method and device for detecting abnormal hyperspectral targets based on band index, which can improve the target detection accuracy of hyperspectral images.

[0006] In a first aspect, an embodiment of the present invention provides a hyperspectral abnormal target detection method based on band index, comprising:

[0007] Calculating the band index of the band according to the image information of the original hyperspectral image; wherein the band index is used to characterize the representative size of the band;

[0008] Performing dimensionality reduction processing on the original hyperspectral image according to the band index to obtain a new hyperspectral image;

[0009] The RX algorithm is used to perform abnormal target detection on the new hyperspectral image to obtain a detection result.

[0010] In a second aspect, an embodiment of the present invention further provides a hyperspectral abnormal target detection device based on band index, comprising:

[0011] A calculation module, used to calculate the band index of the band according to the image information of the original hyperspectral image; wherein the band index is used to characterize the representative size of the band;

[0012] A processing module, used for performing dimensionality reduction processing on the original hyperspectral image according to the band index to obtain a new hyperspectral image;

[0013] The detection module is used to perform abnormal target detection on the new hyperspectral image using the RX algorithm to obtain a detection result.

[0014] In a third aspect, an embodiment of the present invention further provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the method described in any embodiment of this specification is implemented.

[0015] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon, which, when executed in a computer, enables the computer to execute the method described in any embodiment of this specification.

[0016] In a fifth aspect, an embodiment of the present invention further provides a computer program product, including a computer program, which implements the method described in any embodiment of the specification when executed by a processor.

[0017] The embodiment of the present invention provides a method and device for detecting abnormal targets in a hyperspectral image based on band index, which calculates the band index of the original hyperspectral image, arranges the band indexes from large to small, selects the first n bands as the optimal bands, and then forms a new data cube with these bands, performs spectral dimension reduction to obtain a new hyperspectral image, and finally uses the RX algorithm to detect abnormal targets in the new hyperspectral image to obtain abnormal detection results. This method effectively solves the problem of low target detection accuracy and slow calculation speed due to large redundant information between hyperspectral image bands. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0019] Figure 1 It is a flow chart of a method for detecting abnormal hyperspectral targets based on band index provided by an embodiment of the present invention;

[0020] Figure 2 It is a schematic diagram of the AVIRIS hyperspectral image experimental area provided by one embodiment of the present invention;

[0021] Figure 3 It is a schematic diagram of the dimensionality reduction result of the experimental area provided by an embodiment of the present invention;

[0022] Figure 4 It is a schematic diagram of the results of the RX anomaly detection algorithm provided by an embodiment of the present invention;

[0023] Figure 5 is a hardware architecture diagram of an electronic device provided by an embodiment of the present invention;

[0024] Figure 6 It is a structural diagram of a hyperspectral abnormal target detection device based on band index provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0025] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0026] As mentioned above, due to the large redundant information between the bands of hyperspectral images, directly using the RX algorithm to process hyperspectral images will produce a high false alarm probability, resulting in low target detection accuracy and slow calculation speed.

[0027] Based on this, the concept of the present invention is to calculate the band index of the hyperspectral image band, screen the band according to the band index, obtain a new image composed of the optimal band, and then perform target detection on the new image, thereby improving the detection effect and efficiency.

[0028] The specific implementation of the above concept is described below.

[0029] Please refer to Figure 1 The embodiment of the present invention provides a method for detecting abnormal hyperspectral targets based on band index, the method comprising:

[0030] Step 100, calculating the band index of the band according to the image information of the original hyperspectral image; wherein the band index is used to characterize the representative size of the band;

[0031] Step 102, performing dimensionality reduction processing on the original hyperspectral image according to the band index to obtain a new hyperspectral image;

[0032] Step 104: Detect abnormal targets on the new hyperspectral image using the RX algorithm to obtain a detection result.

[0033] In the embodiment of the present invention, the band index of the original hyperspectral image is calculated, the band indexes are arranged from large to small, the first n bands are selected as the optimal bands, and then these bands are combined into a new data cube, and a new hyperspectral image is obtained after spectral dimension reduction, and finally the RX algorithm is used to detect abnormal targets on the new hyperspectral image to obtain abnormal detection results. This method effectively solves the problem of low target detection accuracy and slow calculation speed due to large redundant information between hyperspectral image bands.

[0034] Described below Figure 1 How the various steps are performed.

[0035] First, with respect to step 100 , the band index of the band is calculated according to the image information of the original hyperspectral image.

[0036] Hyperspectral data has a high spectral resolution and can effectively detect differences in spectral characteristics of objects. However, the increase in spectral resolution leads to information redundancy and increases the difficulty of subsequent data processing. For high-dimensional and massive information, data dimensionality reduction must be performed to improve data processing speed.

[0037] As the number of hyperspectral image bands increases, the band width becomes narrower, making the spectral resolution reach below 5nm. Therefore, the spectral resolution of hyperspectral images is sufficient to distinguish substances with diagnostic spectral characteristics. However, when detecting, identifying or classifying objects, the more bands used, the better. This is because:

[0038] 1) The number of spectral channels cannot be simply equated with the number of information channels. Because there is often a strong correlation between adjacent spectral segments, if all spectral segments are used for identification or classification without analysis, it will not only fail to improve the classification accuracy, but will affect the identification or classification results;

[0039] 2) Selecting too many spectral segments not only increases the amount of calculation and affects the calculation speed, but also requires a large number of training samples, otherwise it is difficult to obtain satisfactory recognition or classification results.

[0040] Therefore, it is necessary to follow certain criteria and select effective band combinations to reduce the difficulty of data processing and improve recognition or classification accuracy.

[0041] The band index is a parameter used to characterize the representativeness of the band. It fully considers the information enrichment of each image and the similarity of adjacent bands. The larger the index, the greater the amount of information in the corresponding image and the more representative it is.

[0042] In the embodiment of the present invention, the band index is calculated by the following formula:

[0043] First, the band index I of the band is calculated based on the standard deviation of the band and the correlation coefficient between the band and the two adjacent bands before and after it. i :

[0044]

[0045] In the formula, σ i is the standard deviation of the ith band; R i-1,i is the correlation coefficient between the i-th band and the i-1-th band; R i,i+1 is the correlation coefficient between the i-th band and the i+1-th band;

[0046] The standard deviation of the band is calculated by the following formula:

[0047]

[0048] Where M and N are the number of row and column pixels of the image respectively; f i (x, y) is the image pixel value of the i-th band; is the average value of image pixels in the i-th band;

[0049] The correlation coefficient is calculated using the following formula:

[0050]

[0051] In the formula, R i,j is the correlation coefficient between the i-th band and the j-th band; E{} is the mathematical expectation; f i is the pixel value of the i-th image; is the pixel average value of the i-th image; f j is the pixel value of the jth image; is the pixel average of the jth image.

[0052] Then, for step 102, the original hyperspectral image is subjected to dimensionality reduction processing according to the band index to obtain a new hyperspectral image.

[0053] In an embodiment of the present invention, the dimensionality reduction process includes the following steps: arranging all the band indexes in descending order; selecting the first n bands as the optimal bands according to a preset band selection standard; and obtaining the new hyperspectral image by forming a data cube composed of the optimal bands.

[0054] In the embodiment of the present invention, the selection criteria of the optimal band are: determining a band with the largest amount of information, a band data correlation less than a first threshold, and a ground object spectrum difference greater than a second threshold as an optimal band.

[0055] Specifically, the characteristics of the large number of hyperspectral data bands and narrow band width determine that the correlation between hyperspectral data bands is large and the information overlap is high. There are three principles for selecting the optimal band: first, the amount of information in the selected band should be the largest; second, the correlation between the band data should be small; third, the spectral response characteristics of the objects to be identified in the study area can make it easiest to distinguish between certain types of objects. The bands with high information content, low correlation, large spectral differences of objects, and good separability are the optimal bands that should be selected.

[0056] With respect to step 104, the RX algorithm is used to perform abnormal target detection on the new hyperspectral image to obtain a detection result.

[0057] The specific detection process is well known to those skilled in the art and will not be described in detail here. The following is an example data to demonstrate the effectiveness of the above method:

[0058] The data selected in the embodiment of the present invention is the 224-band AVIRIS hyperspectral image acquired at the Naval Airfield in San Diego, California, USA, with a wavelength range of 0.37-2.51um, a spatial resolution of 3.5m, and a size of 400*400 pixels. After removing the water vapor absorption band and the low signal-to-noise ratio band, 189 bands are retained. Due to the large amount of data, in order to improve the experimental efficiency, an area of ​​100*100 pixels is intercepted, which contains a large number of aircraft targets to be extracted, such as Figure 2 shown.

[0059] First, the band index I is calculated for the original hyperspectral image. i The indexes of each band are arranged from large to small, the first n bands are selected as the optimal bands, and these bands are combined into a new data cube to obtain a new hyperspectral image. The hyperspectral image obtained in this embodiment has 73 bands.

[0060] For the hyperspectral image after dimensionality reduction, the RX algorithm is used to detect abnormal targets and obtain the abnormal detection results, such as Figure 3 At the same time, the RX algorithm is used to detect abnormal targets directly on the original hyperspectral image, and the abnormal detection results are as follows Figure 4 shown.

[0061] contrast Figure 3 and Figure 4 ,It can be seen that after the band selection and then the anomaly detection, the target detection effect is greatly improved, and the speed and efficiency of the algorithm are also improved.

[0062] In summary, the method provided by the embodiment of the present invention significantly improves the abnormal target detection effect, and solves the problem of low target detection accuracy and slow calculation speed due to large redundant information between hyperspectral image bands.

[0063] like Figure 5 , Figure 6 As shown, the embodiment of the present invention provides a hyperspectral abnormal target detection device based on band index. The device embodiment can be implemented by software, or by hardware or a combination of software and hardware. From the hardware level, Figure 5 As shown, it is a hardware architecture diagram of an electronic device in which a hyperspectral abnormal target detection device based on band index provided by an embodiment of the present invention is located. Figure 5 In addition to the processor, memory, network interface, and non-volatile memory shown, the electronic device in the embodiment may also include other hardware, such as a forwarding chip responsible for processing messages, etc. Taking software implementation as an example, Figure 6 As shown, as a device in a logical sense, the CPU of the electronic device in which it is located reads the corresponding computer program in the non-volatile memory into the memory and runs it. This embodiment provides a hyperspectral abnormal target detection device based on band index, including:

[0064] The calculation module 600 is used to calculate the band index of the band according to the image information of the original hyperspectral image; wherein the band index is used to characterize the representative size of the band;

[0065] A processing module 602 is used to perform dimensionality reduction processing on the original hyperspectral image according to the band index to obtain a new hyperspectral image;

[0066] The detection module 604 is used to perform abnormal target detection on the new hyperspectral image using the RX algorithm to obtain a detection result.

[0067] In the embodiment of the present invention, when the calculation module 600 calculates the band index of the band according to the image information of the original hyperspectral image, it is specifically used to perform the following operations:

[0068] The band index I of the band is calculated based on the standard deviation of the band and the correlation coefficient between the band and the two adjacent bands before and after it. i :

[0069]

[0070] In the formula, σ i is the standard deviation of the ith band; R i-1,i is the correlation coefficient between the i-th band and the i-1-th band; R i,i+1 is the correlation coefficient between the i-th band and the i+1-th band;

[0071] In the embodiment of the present invention, the standard deviation of the band is calculated by the following formula:

[0072]

[0073] Where M and N are the number of row and column pixels of the image respectively; f i (x, y) is the image pixel value of the i-th band; is the average value of image pixels in the i-th band;

[0074] In the embodiment of the present invention, the correlation coefficient is calculated by the following formula:

[0075]

[0076] In the formula, R i,j is the correlation coefficient between the i-th band and the j-th band; E{} is the mathematical expectation; f i is the pixel value of the i-th image; is the pixel average value of the i-th image; f j is the pixel value of the jth image; is the pixel average of the jth image.

[0077] In the embodiment of the present invention, when the processing module 602 performs dimensionality reduction processing on the original hyperspectral image according to the band index to obtain a new hyperspectral image, it is specifically used to perform the following operations: arranging all the band indexes in order from large to small; selecting the first n bands as the optimal bands according to a preset band selection standard; and obtaining the new hyperspectral image by forming a data cube composed of the optimal bands.

[0078] In an embodiment of the present invention, the first n bands are selected as optimal bands according to a preset band selection standard, including: determining a band with the largest amount of information, a band data piece correlation less than a first threshold, and a ground object spectral difference greater than a second threshold as an optimal band.

[0079] It is understood that the structure illustrated in the embodiment of the present invention does not constitute a specific limitation on a hyperspectral abnormal target detection device based on a band index. In other embodiments of the present invention, a hyperspectral abnormal target detection device based on a band index may include more or fewer components than shown in the figure, or combine certain components, or split certain components, or arrange the components differently. The illustrated components may be implemented in hardware, software, or a combination of software and hardware.

[0080] The information interaction, execution process and other contents between the modules in the above-mentioned device are based on the same concept as the embodiment of the method of the present invention. For specific contents, please refer to the description in the embodiment of the method of the present invention, and no further description is given here.

[0081] An embodiment of the present invention further provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, a hyperspectral abnormal target detection method based on band index in any embodiment of the present invention is implemented.

[0082] An embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the processor executes a hyperspectral abnormal target detection method based on band index in any embodiment of the present invention.

[0083] An embodiment of the present application also provides a computer program product, which includes a computer program. A processor of a computer device reads the computer program from a computer-readable storage medium, and the processor executes the computer program, so that the computer device executes any of the band index-based hyperspectral abnormal target detection methods described in the above embodiments.

[0084] Specifically, a system or device equipped with a storage medium can be provided, on which software program code that implements the functions of any of the above-mentioned embodiments is stored, and a computer (or CPU or MPU) of the system or device can be enabled to read and execute the program code stored in the storage medium.

[0085] In this case, the program code itself read from the storage medium can realize the function of any one of the above-mentioned embodiments, and thus the program code and the storage medium storing the program code constitute a part of the present invention.

[0086] The storage medium embodiments for providing the program code include a floppy disk, a hard disk, a magneto-optical disk, an optical disk (such as CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RAM, DVD-RW, DVD+RW), a magnetic tape, a non-volatile memory card, and a ROM. Alternatively, the program code can be downloaded from a server computer by a communication network.

[0087] In addition, it should be clear that the functions of any of the above embodiments can be implemented not only by executing the program code read by the computer, but also by enabling an operating system operating on the computer to complete part or all of the actual operations based on instructions from the program code.

[0088] In addition, it can be understood that the program code read from the storage medium is written to a memory provided in an expansion board inserted into the computer or to a memory provided in an expansion module connected to the computer, and then based on the instructions of the program code, a CPU installed on the expansion board or expansion module is enabled to perform part or all of the actual operations, thereby realizing the functions of any of the above-mentioned embodiments.

[0089] It should be noted that, in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the statement "comprise a ..." do not exclude the presence of other identical factors in the process, method, article or device including the elements.

[0090] A person of ordinary skill in the art can understand that all or part of the steps of implementing the above method embodiments can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiments; and the aforementioned storage medium includes: ROM, RAM, magnetic disk or optical disk, etc., various media that can store program codes.

[0091] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A hyperspectral abnormal target detection method based on band index, characterized in that: include: Calculating the band index of the band according to the image information of the original hyperspectral image; wherein the band index is used to characterize the representative size of the band; Performing dimensionality reduction processing on the original hyperspectral image according to the band index to obtain a new hyperspectral image; The RX algorithm is used to perform abnormal target detection on the new hyperspectral image to obtain a detection result.

2. The method according to claim 1, characterized in that The step of calculating the band index of each band according to the image information of the original hyperspectral image includes: The band index I of the band is calculated based on the standard deviation of the band and the correlation coefficient between the band and the two adjacent bands before and after it. i : In the formula, σ i is the standard deviation of the ith band; R i-1,i is the correlation coefficient between the i-th band and the i-1-th band; R i,i+1 is the correlation coefficient between the i-th band and the i+1-th band.

3. The method according to claim 2, characterized in that The standard deviation of the band is calculated by the following formula: Where M and N are the number of row and column pixels of the image respectively; f i (x, y) is the image pixel value of the i-th band; is the average value of image pixels in the i-th band.

4. The method according to claim 2, characterized in that: The correlation coefficient is calculated by the following formula: In the formula, R i,j is the correlation coefficient between the i-th band and the j-th band; E{} is the mathematical expectation; f i is the pixel value of the i-th image; is the pixel average value of the i-th image; f j is the pixel value of the jth image; is the pixel average of the jth image.

5. The method according to claim 1, characterized in that The step of performing dimensionality reduction processing on the original hyperspectral image according to the band index to obtain a new hyperspectral image includes: Arrange all the band indices in order from largest to smallest; Select the first n bands as the optimal bands according to the preset band selection criteria; The data cube composed of the optimal bands is obtained to obtain the new hyperspectral image.

6. The method according to claim 5, characterized in that The step of selecting the first n bands as optimal bands according to a preset band selection standard includes: A band with the largest amount of information, a band data piece correlation less than a first threshold, and a ground object spectrum difference greater than a second threshold is determined as an optimal band.

7. A hyperspectral abnormal target detection device based on band index, characterized in that: include: A calculation module, used to calculate the band index of the band according to the image information of the original hyperspectral image; wherein the band index is used to characterize the representative size of the band; A processing module, used for performing dimensionality reduction processing on the original hyperspectral image according to the band index to obtain a new hyperspectral image; The detection module is used to perform abnormal target detection on the new hyperspectral image using the RX algorithm to obtain a detection result.

8. An electronic device, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the method according to any one of claims 1 to 6 is implemented.

9. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed in a computer, the computer is caused to execute the method according to any one of claims 1 to 6.

10. A computer-readable storage medium, characterized in that: The method comprises a computer program, wherein when the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images