Spaceborne SAR time-series image change detection method and device

Abstract

The present invention discloses a method and device for detecting changes in spaceborne SAR time-series images. The method comprises: registering a spaceborne SAR time-series image sequence to obtain a registered image sequence; extracting a background image from the registered image sequence; subtracting the background image from the registered time-series image sequence to generate a foreground image sequence; adding a binary mask to the foreground image sequence to obtain a region to be detected; and detecting target change pixels in the region to be detected to obtain change detection results for the spaceborne SAR time-series images. The present invention can perform change detection on spaceborne SAR time-series images with high accuracy and reliability.

Description

Technical Field

[0001] The present invention relates to the technical field of image change detection, and in particular to a method and device for detecting changes in spaceborne SAR time-series images. Background Art

[0002] This section is intended to provide a background or context to the embodiments of the invention that are recited in the claims. No statement herein is admitted to be prior art by virtue of its inclusion in this section.

[0003] Traditional change detection methods for spaceborne SAR imagery are primarily based on pixel-level differences, which detect changes by comparing pixel-level differences in SAR images taken over the same area at different points in time. While simple and easy to implement, this method is subject to interference from noise and clutter, and its accuracy and reliability are affected by false positives and false negatives. Furthermore, due to the large amount of information and complex nature of SAR images, pixel-level difference comparison methods often struggle to pinpoint the precise location and shape of changed areas, limiting the accuracy and scope of change detection. Summary of the Invention

[0004] An embodiment of the present invention provides a method for detecting changes in spaceborne SAR time-series images, which is used to detect changes in spaceborne SAR time-series images with high accuracy and high reliability. The method includes:

[0005] Register the spaceborne SAR time series image sequence to obtain the registered image sequence;

[0006] Extracting the background image from the registered image sequence;

[0007] Subtract the background image from the registered temporal image sequence to generate a foreground image sequence;

[0008] Add a binary mask to the foreground image sequence to obtain the area to be detected;

[0009] The target change pixels are detected from the area to be detected, and the change detection results of the spaceborne SAR time series image are obtained.

[0010] An embodiment of the present invention further provides a spaceborne SAR time-series image change detection device for detecting changes in spaceborne SAR time-series images, which has high accuracy and high reliability. The device includes:

[0011] A registration module is used to register the spaceborne SAR time-series image sequence to obtain a registered image sequence;

[0012] A background image extraction module, used to extract the background image from the registered image sequence;

[0013] a foreground image sequence generation module, configured to generate a foreground image sequence by subtracting a background image from the registered temporal image sequence;

[0014] The module for obtaining the region to be detected is used to add a binary mask to the foreground image sequence to obtain the region to be detected;

[0015] The target change pixel point detection module is used to detect target change pixels from the area to be detected and obtain change detection results of the spaceborne SAR time series image.

[0016] An embodiment of the present invention further provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned method for detecting changes in spaceborne SAR time-series images when executing the computer program.

[0017] An embodiment of the present invention further provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the computer program implements the above-mentioned method for detecting changes in spaceborne SAR time-series images.

[0018] An embodiment of the present invention further provides a computer program product, which includes a computer program. When the computer program is executed by a processor, the computer program implements the above-mentioned spaceborne SAR time-series image change detection method.

[0019] In an embodiment of the present invention, a spaceborne SAR time-series image sequence is registered to obtain a registered image sequence; a background image is extracted from the registered image sequence; the background image is subtracted from the registered time-series image sequence to generate a foreground image sequence; a binary mask is added to the foreground image sequence to obtain a region to be detected; target change pixels are detected from the region to be detected to obtain a change detection result for the spaceborne SAR time-series image. Compared with the prior art technical solution for spaceborne SAR image change detection based on the pixel difference method, the embodiment of the present invention applies logarithmic background difference to spaceborne SAR time-series image changes, that is, extracting the background image from the registered image sequence, subtracting the background image from the registered time-series image sequence to generate a foreground image sequence, which is then used for subsequent detection. This solves the problems of low change detection efficiency and susceptibility to clutter and noise interference, and achieves high detection accuracy and reliability. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative work. In the drawings:

[0021] Figure 1 Flowchart of a method for detecting changes in spaceborne SAR time-series images according to an embodiment of the present invention;

[0022] Figure 2 This is the principle diagram of change detection for spaceborne SAR time series images;

[0023] Figure 3 This is a flow chart of registering a spaceborne SAR time-series image sequence using the SAR-SIFT algorithm in an embodiment of the present invention;

[0024] Figure 4 Schematic diagram of a spaceborne SAR time-series image change detection device according to an embodiment of the present invention;

[0025] Figure 5 Schematic diagram of a computer device in an embodiment of the present invention. DETAILED DESCRIPTION

[0026] To make the purpose, technical solutions and advantages of the embodiments of the present invention more clear, the embodiments of the present invention are further described in detail below with reference to the accompanying drawings. Here, the exemplary embodiments of the present invention and their descriptions are used to explain the present invention, but are not intended to limit the present invention.

[0027] The inventors discovered that the logarithmic background difference method can be applied to SAR image change detection. This method extracts the background from multiple SAR images in a time-series sequence, then subtracts the background image from the current image to determine the pixels representing the changed area, thereby performing change detection. Compared to the pixel difference method, the logarithmic background difference method not only eliminates the influence of clutter and noise but also performs multi-scale image processing, enabling more accurate detection of the location and shape of changed areas.

[0028] Figure 1 Flowchart of a method for detecting changes in spaceborne SAR time-series images according to an embodiment of the present invention. Figure 2 The schematic diagram of the spaceborne SAR time series image change detection method is as follows:

[0029] Step 101, registering a spaceborne SAR time-series image sequence to obtain a registered image sequence;

[0030] Step 102, extracting the background image from the registered image sequence;

[0031] Step 103, subtracting the background image from the registered time-series image sequence to generate a foreground image sequence;

[0032] Step 104: Add a binary mask to the foreground image sequence to obtain the area to be detected;

[0033] Step 105 : Detect target changed pixels from the area to be detected, and obtain change detection results of the spaceborne SAR time series image.

[0034] Compared with the prior art technical solution for detecting changes in spaceborne SAR images based on the pixel difference method, the embodiment of the present invention applies logarithmic background difference to changes in spaceborne SAR time-series images, solving the problems of low change detection efficiency and susceptibility to clutter and noise interference, and achieving high detection accuracy and reliability.

[0035] In step 101, a spaceborne SAR time-series image sequence is registered to obtain a registered image sequence;

[0036] When a spaceborne SAR platform repeatedly observes the same scene, it is affected by different time and orbital offsets, which results in large geometric positioning errors in the images, making it impossible to directly perform background extraction and change detection. Therefore, configuration is required first.

[0037] In this embodiment of the present invention, the SAR-SIFT algorithm is used to register a spaceborne SAR time-series image sequence. The spaceborne SAR time-series image sequence consists of n images, each of which has a size of M × N pixels. It is represented by a three-dimensional array f{I} of size n × M × N, where the kth layer represents the kth image, and each layer has M × N pixels. This can be expressed as:

[0038] f{I}={I(i,j,k),i=1,2,...M;j=1,2,...N;k=1,2,...n} (1)

[0039] Where I(i,j,k) represents the pixel located at the i-th row and j-th column in the k-th image.

[0040] Image registration is the process of spatially aligning two images in a scene so that corresponding points in the two images have the same coordinates. This is achieved by finding a transformation matrix that transforms the image to be registered into the same coordinate system as the reference image, thereby achieving spatial alignment between the image to be registered and the reference image.

[0041] See also Figure 3 This is a flow chart of registering a spaceborne SAR time-series image sequence using the SAR-SIFT algorithm in an embodiment of the present invention. Registering a spaceborne SAR time-series image sequence using the SAR-SIFT algorithm to obtain a registered image sequence includes:

[0042] Step 301, extracting characteristic pixel points of the spaceborne SAR time series image and the reference image respectively;

[0043] Step 302: matching the characteristic pixel points of the spaceborne SAR time-series image with the characteristic pixel points of the reference image to obtain a matching pair of characteristic pixel points;

[0044] In one embodiment, matching characteristic pixel points of a spaceborne SAR time-series image with characteristic pixel points of a reference image to obtain a matching pair of characteristic pixel points includes:

[0045] Using the nearest neighbor distance ratio (NNDR), feature pixels are matched to obtain multiple feature pixel matching pairs. The specific steps are as follows: selecting two feature pixels with the closest Euclidean distance, determining whether the Euclidean distance between the two feature pixels meets a preset threshold; if so, determining that the two feature pixels are a feature pixel matching pair;

[0046] The calculation formula of Euclidean distance is:

[0047]

[0048] Where d is the Euclidean distance.

[0049] Afterwards, the RANSAC algorithm is used to remove the mismatched feature pixel matching pairs and obtain the correct feature pixel matching pairs.

[0050] The basic assumption of the RANSAC algorithm is that the sample contains both correct data (inliers, data that can be described by the model) and abnormal data (outliers, data that deviates significantly from the normal range and cannot be fit into the mathematical model). In other words, the dataset contains noise. This abnormal data may be caused by incorrect measurements, incorrect assumptions, or incorrect calculations. RANSAC also assumes that, given a set of correct data, there is a method to calculate the model parameters that fit this data.

[0051] Step 303, using the feature pixel matching pairs, calculate the transformation model parameters to obtain the transformation matrix;

[0052] Use I1(x,y) and I2(x',y') to represent the pixel value at (x,y) in the reference image and the image to be registered, respectively. The image registration can be expressed by the following mathematical formula:

[0053] I2(x',y')=f[I1(x,y)] (3)

[0054] Where f is the coordinate transformation process. The SAR-SIFT algorithm used in the embodiments of the present invention is a feature-based image registration algorithm. This algorithm matches the same-name features in the reference image with the image to be registered to obtain the parameters of the transformation model. Finally, the registered image is obtained through resampling and interpolation.

[0055] Step 304: resample the spaceborne SAR time-series image sequence according to the transformation matrix to obtain a registered image sequence. During resampling, the position and size of the spaceborne SAR time-series image sequence need to be adjusted.

[0056] Step 102: extract the background image from the registered image sequence. In specific implementation, a median filter may be applied to the registered image sequence along the azimuth and time dimensions to obtain the background image. The principle of applying the median filter is as follows:

[0057] The grayscale values ​​of each pixel in multiple images are sorted and the median value is taken as the grayscale value of the pixel in the background image to obtain the background image. In practical applications, different median filters of different sizes can be selected for experimental comparison.

[0058]

[0059] Among them, B(i,j) represents the grayscale value of the pixel located in the i-th row and j-th column in the background image. Represents the image sequence after registration, and median represents the operation of taking the median.

[0060] In step 103, the background image is subtracted from the registered temporal image sequence to generate a foreground image sequence. The formula is as follows:

[0061]

[0062] Among them, F(i,j,k) represents the grayscale value of the pixel located in the kth image, i-th row, and j-th column in the foreground image sequence.

[0063] In step 104, a binary mask is added to the foreground image sequence to obtain the area to be detected;

[0064] A binary mask is added to the foreground image sequence to exclude other interfering areas, leaving only the area to be detected. Specifically, a binary mask needs to be applied to each image in the foreground image sequence.

[0065]

[0066] in, Represents the pixel at the kth image, ith row, and jth column in the foreground image sequence after the binary mask processing. M(i,j) represents the pixel at the ith row and jth column in the binary mask image (can only be 0 or 1). When multiplying pixel by pixel, a pixel with a value of 0 in the binary mask will cause the corresponding pixel in the foreground image sequence to become 0, while a pixel with a value of 1 will have no effect.

[0067] Step 105 : Detect target changed pixels from the area to be detected, and obtain change detection results of the spaceborne SAR time series image.

[0068] In one embodiment, detecting target change pixels in the area to be detected and obtaining change detection results of spaceborne SAR time-series images includes:

[0069] Apply the CFAR algorithm to detect target change pixels from the area to be detected;

[0070] Count the number of target changed pixels;

[0071] Determine the target changed pixels and quantity as the change detection result.

[0072] Among them, the CFAR algorithm first processes the input noise and then determines a threshold, and compares this threshold with the input signal. If the input signal exceeds this threshold, it is judged as a target, otherwise, it is judged as no target.

[0073] The embodiment of the present invention further provides a spaceborne SAR time-series image change detection device, the principle of which is similar to the spaceborne SAR time-series image change detection method, and will not be described in detail here.

[0074] Figure 4 Schematic diagram of a spaceborne SAR time-series image change detection device according to an embodiment of the present invention, comprising:

[0075] A registration module 401 is used to register a spaceborne SAR time-series image sequence to obtain a registered image sequence;

[0076] Background image extraction module 402, used to extract the background image from the registered image sequence;

[0077] A foreground image sequence generating module 403 is configured to generate a foreground image sequence by subtracting the background image from the registered time-series image sequence;

[0078] The module 404 for obtaining the region to be detected is used to add a binary mask to the foreground image sequence to obtain the region to be detected;

[0079] The target change pixel point detection module 405 is used to detect target change pixel points from the area to be detected and obtain change detection results of the spaceborne SAR time series image.

[0080] In one embodiment, the registration module is specifically configured to:

[0081] Extract the characteristic pixel points of the spaceborne SAR time series image and the characteristic pixel points of the reference image respectively;

[0082] Match the characteristic pixel points of the spaceborne SAR time series image with the characteristic pixel points of the reference image to obtain the characteristic pixel point matching pairs;

[0083] Using the matching pairs of feature pixels, the transformation model parameters are calculated to obtain the transformation matrix;

[0084] According to the transformation matrix, the spaceborne SAR time-series image sequence is resampled to obtain a registered image sequence. In one embodiment, the registration module is specifically used to:

[0085] Using the nearest neighbor ratio, the characteristic pixels of the spaceborne SAR time series image are matched with the characteristic pixels of the reference image to obtain multiple matching pairs of characteristic pixels;

[0086] The RANSAC algorithm is used to remove mismatched feature pixel matching pairs and obtain correct feature pixel matching pairs.

[0087] In one embodiment, the registration module is specifically configured to:

[0088] Select two feature pixels with the closest Euclidean distance, and determine whether the Euclidean distance between the two feature pixels meets a preset threshold;

[0089] If so, the two feature pixels are determined to be a feature pixel matching pair.

[0090] In one embodiment, the background image extraction module is specifically configured to:

[0091] A median filter is applied to the registered image sequence along the azimuth-time dimension to obtain the background image.

[0092] In one embodiment, the target change pixel detection module is specifically used to:

[0093] Apply the CFAR algorithm to detect target change pixels from the area to be detected;

[0094] Count the number of target changed pixels;

[0095] Determine the target changed pixels and quantity as the change detection result.

[0096] The beneficial effects achieved by the method and device described in the embodiments of the present invention are as follows:

[0097] First, the logarithmic background difference algorithm is applied to the field of change detection in spaceborne SAR images. Compared with the method based on pixel difference, it solves the problems of low change detection efficiency and susceptibility to clutter and noise interference.

[0098] Second, the logarithmic background difference method is combined with the SAR-SIFT algorithm and improved to make it suitable for the needs of change detection in spaceborne SAR time series images.

[0099] An embodiment of the present invention further provides a computer device, Figure 5 Schematic diagram of a computer device in an embodiment of the present invention. The computer device 500 includes a memory 510, a processor 520, and a computer program 530 stored in the memory 510 and executable on the processor 520. When the processor 520 executes the computer program 530, the above-mentioned spaceborne SAR time-series image change detection method is implemented.

[0100] An embodiment of the present invention further provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the computer program implements the above-mentioned method for detecting changes in spaceborne SAR time-series images.

[0101] An embodiment of the present invention further provides a computer program product, which includes a computer program. When the computer program is executed by a processor, the computer program implements the above-mentioned spaceborne SAR time-series image change detection method.

[0102] It will be understood by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0103] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0104] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0105] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0106] The specific embodiments described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for detecting changes in spaceborne SAR time series images, characterized in that: include: Register the spaceborne SAR time series image sequence to obtain the registered image sequence; Registering a spaceborne SAR time-series image sequence to obtain a registered image sequence, including: extracting characteristic pixels of the spaceborne SAR time-series image and characteristic pixels of a reference image respectively; matching the characteristic pixels of the spaceborne SAR time-series image with the characteristic pixels of the reference image to obtain characteristic pixel matching pairs; calculating transformation model parameters using the characteristic pixel matching pairs to obtain a transformation matrix; resampling the spaceborne SAR time-series image sequence according to the transformation matrix to obtain a registered image sequence; matching the characteristic pixels of the spaceborne SAR time-series image with the characteristic pixels of the reference image to obtain characteristic pixel matching pairs, including: matching the characteristic pixels of the spaceborne SAR time-series image with the characteristic pixels of the reference image using a nearest neighbor ratio to obtain multiple characteristic pixel matching pairs; and removing mismatched characteristic pixel matching pairs using a RANSAC algorithm to obtain correct characteristic pixel matching pairs; Extracting the background image from the registered image sequence; Subtract the background image from the registered temporal image sequence to generate a foreground image sequence; Add a binary mask to the foreground image sequence to obtain the area to be detected; Detecting target change pixels from the area to be detected to obtain change detection results of spaceborne SAR time series images; detecting target change pixels from the area to be detected to obtain change detection results of spaceborne SAR time series images, including: applying the CFAR algorithm to detect target change pixels from the area to be detected; counting the number of target change pixels; and determining the target change pixels and the number as the change detection results.

2. The method according to claim 1, wherein Using the nearest neighbor ratio, the characteristic pixels of the spaceborne SAR time series image are matched with the characteristic pixels of the reference image to obtain multiple characteristic pixel matching pairs, including: Select two feature pixels with the closest Euclidean distance, and determine whether the Euclidean distance between the two feature pixels meets a preset threshold; If so, the two feature pixels are determined to be a feature pixel matching pair.

3. The method according to claim 1, wherein Extract the background image from the registered image sequence, including: A median filter is applied to the registered image sequence along the azimuth-time dimension to obtain the background image.

4. A spaceborne SAR time series image change detection device, characterized in that: include: A registration module is used to register the spaceborne SAR time-series image sequence to obtain a registered image sequence; Registering a spaceborne SAR time-series image sequence to obtain a registered image sequence, including: extracting characteristic pixels of the spaceborne SAR time-series image and characteristic pixels of a reference image respectively; matching the characteristic pixels of the spaceborne SAR time-series image with the characteristic pixels of the reference image to obtain characteristic pixel matching pairs; calculating transformation model parameters using the characteristic pixel matching pairs to obtain a transformation matrix; resampling the spaceborne SAR time-series image sequence according to the transformation matrix to obtain a registered image sequence; matching the characteristic pixels of the spaceborne SAR time-series image with the characteristic pixels of the reference image to obtain characteristic pixel matching pairs, including: matching the characteristic pixels of the spaceborne SAR time-series image with the characteristic pixels of the reference image using a nearest neighbor ratio to obtain multiple characteristic pixel matching pairs; and removing mismatched characteristic pixel matching pairs using a RANSAC algorithm to obtain correct characteristic pixel matching pairs; A background image extraction module, used to extract the background image from the registered image sequence; a foreground image sequence generation module, configured to generate a foreground image sequence by subtracting a background image from the registered temporal image sequence; The module for obtaining the region to be detected is used to add a binary mask to the foreground image sequence to obtain the region to be detected; The target change pixel detection module is used to detect target change pixels from the area to be detected and obtain change detection results of the spaceborne SAR time series image; detecting target change pixels from the area to be detected and obtaining change detection results of the spaceborne SAR time series image includes: applying the CFAR algorithm to detect target change pixels from the area to be detected; counting the number of target change pixels; and determining the target change pixels and their number as the change detection results.

5. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the method according to any one of claims 1 to 3 is implemented.

6. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 3 is implemented.

7. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 3 is implemented.

Images