Imaging deformation correction method, device and system based on bidirectional push-broom velocity estimation

Abstract

The present application relates to remote sensing image processing technical field, especially in kind of based on two-way push scanning speed estimation's imaging deformation correction method, device and system. The method constructs the imaging deformation parameter model by analyzing the reason of the non-cooperative target push scanning imaging deformation in the air; based on the imaging mechanism of positive and negative two-way push scanning, the difference of the relative motion of the satellite imaging platform and the target under the two-way push scanning imaging is analyzed, and the speed of the non-cooperative target in the air is estimated; based on the speed estimation result, combined with the imaging parameters of the satellite load platform, the imaging deformation model parameters are solved, the spatial resampling frequency of the non-cooperative target in the air is determined, and the target morphology reconstruction is realized. The method is a deformation fast correction method which does not depend on the prior information of the target and is completely based on data driving, estimates the motion parameters and corrects the deformation adaptively through the internal difference of the two-way push scanning data, and meets the timeliness requirement of remote sensing data processing.

Description

Technical Field

[0001] This invention relates to the field of remote sensing image processing technology, and in particular to an imaging deformation correction method, apparatus and system based on bidirectional push-broom velocity estimation. Background Technology

[0002] Satellite hyperspectral imaging technology, especially pushbroom imaging mode, is widely used in target reconnaissance and identification, as well as Earth observation, due to its advantages such as wide imaging coverage, rich spectral information, and integrated image and spectrum. In recent years, its application scope has expanded to the detection, identification, and characteristic measurement of non-cooperative aerial targets.

[0003] However, when a satellite hyperspectral payload performs pushbroom imaging of a non-cooperative aerial target, a complex relative motion exists between the imaging parameters (such as spatial resolution and pushbroom imaging velocity) and the target's relative motion parameters (such as target velocity). This results in a mismatch between the effective scan length and the imaging field of view of the target in the pushbroom direction, causing geometric deformation of the target image in the pushbroom direction, resulting in stretching or compression (see [reference]). Figure 1 This deformation directly affects the accurate extraction of key characteristics such as the target's radiation area and radiation intensity, thereby affecting the accuracy of the target's radiation characteristic analysis.

[0004] Furthermore, for non-cooperative aerial targets, prior information such as their geometry, speed, and heading is often difficult to obtain, making traditional deformation correction methods that rely on prior target information unfeasible. Currently, there is a lack of effective correction methods for target imaging deformation caused by pushbroom imaging of non-cooperative aerial targets using satellite hyperspectral payloads.

[0005] Therefore, there is an urgent need for a technical solution that can quickly and accurately correct such imaging deformations without relying on prior information about the target. Summary of the Invention

[0006] The purpose of this invention is to provide an imaging deformation correction method, device and system based on bidirectional pushbroom velocity estimation, to construct an imaging deformation parameter model for non-cooperative aerial targets, and to solve the target deformation problem caused by the mismatch between imaging parameters and target relative motion state parameters.

[0007] To achieve the above objectives, in a first aspect, the present invention provides an imaging deformation correction method based on bidirectional push-broom velocity estimation, characterized by comprising the following steps: S1. Based on the pushbroom imaging mechanism of satellite hyperspectral payload, establish a target scaling factor model to characterize the target imaging deformation caused by the relative motion between the satellite platform and the non-cooperative target in the air along the pushbroom direction. S2. Acquire the forward pushbroom imaging data and backward pushbroom imaging data of the non-cooperative aerial target, as well as the imaging parameters of the satellite platform; extract the first pixel geometric dimension of the non-cooperative aerial target in the pushbroom direction from the forward and backward pushbroom imaging data, respectively. Second pixel geometry ; S3. Based on the imaging parameters of the satellite platform, the geometric dimensions of the first pixel and the second pixel, analyze the relative motion relationship between the satellite platform and the non-cooperative aerial target under forward and reverse pushbroom conditions, and solve for the velocity of the non-cooperative aerial target along the pushbroom direction. ; S4, the speed Substituting into the target scaling factor model, the target scaling factor is calculated. Based on the target scaling factor The infrared hyperspectral image data of the mask area where the non-cooperative aerial target is located is spatially resampled along the push-broom direction to correct imaging distortion.

[0008] Optionally, in step S1, the target scaling factor model is: in, , ; In the formula, Zoom ratio for pushbroom imaging of non-cooperative aerial targets. The equivalent spatial resolution of the target along the push-broom direction, The field-of-view spatial resolution is defined by the vertical push-broom direction. The relative velocity between the satellite imaging platform and the non-cooperative aerial target along the push-broom direction. For the imaging period of the satellite hyperspectral imaging payload, The distance between the satellite's hyperspectral imaging payload and a non-cooperative target in the air. This represents the instantaneous field of view of the satellite's hyperspectral payload.

[0009] Optionally, in step S2, the non-cooperative aerial target is extracted from the forward and reverse pushbroom imaging data using an image segmentation method to generate a target mask image, and the geometric dimensions of the first pixel are obtained through connected component analysis. Second pixel geometry .

[0010] Optionally, the image segmentation method is a threshold segmentation method or an edge segmentation method.

[0011] Optionally, in step S3, the relative motion relationship satisfies: During forward pushbrooming, the relative velocity between the satellite platform and the non-cooperative aerial target along the pushbroom direction. ; During reverse pushbroom operation, the relative velocity between the satellite platform and the non-cooperative aerial target along the pushbroom direction. ; in, The scanning speed of the satellite hyperspectral payload platform along the push-broom direction. The velocity of the non-cooperative aerial target along the push-broom direction.

[0012] Optionally, the differences in imaging effects of non-cooperative aerial targets under bidirectional pushbroom scanning are analyzed, and the target pixel sizes exhibit the following relationship: Based on the geometric pixel extraction results of the non-cooperative aerial target under bidirectional pushbroom imaging, and combined with the established geometric pixel relative relationships of the non-cooperative aerial target under bidirectional pushbroom imaging, the velocity of the non-cooperative aerial target along the pushbroom direction is calculated using the following formula. : .

[0013] Optionally, the target pushbroom imaging scaling ratio can be calculated using the following formula. : .

[0014] Optionally, in step S4, the spatial resampling employs nearest neighbor interpolation or linear interpolation.

[0015] Secondly, the present invention also provides an imaging deformation correction device based on bidirectional push-broom velocity estimation, comprising: The model building module is used to build an imaging deformation parameter model: based on the pushbroom imaging mechanism of the satellite hyperspectral payload, a target scaling ratio model is established to characterize the target imaging deformation caused by the relative motion between the satellite platform and the non-cooperative target in the air along the pushbroom direction. The data processing module is used for imaging parameter analysis and target two-way pushbroom data processing: acquiring the forward and reverse pushbroom imaging data of the non-cooperative aerial target, as well as the imaging parameters of the satellite platform; and extracting the first pixel geometric dimension of the non-cooperative aerial target in the pushbroom direction from the forward and reverse pushbroom imaging data, respectively. Second pixel geometry ; The velocity estimation module is used for target velocity estimation: based on the imaging parameters of the satellite platform, the geometry of the first pixel, and the geometry of the second pixel, it analyzes the relative motion between the satellite platform and the non-cooperative aerial target under forward and reverse pushbroom conditions, and solves for the velocity of the non-cooperative aerial target along the pushbroom direction. ; The correction execution module is used for deformation correction: [The sentence fragment "the speed" appears to be incomplete and lacks context. It's unclear what "the speed" refers to.] Substituting into the target scaling factor model, the target scaling factor is calculated. Based on the target scaling factor The infrared hyperspectral image data of the mask area where the non-cooperative aerial target is located is spatially resampled along the push-broom direction to correct imaging distortion.

[0016] Thirdly, the present invention also provides an imaging deformation correction system based on bidirectional push-broom velocity estimation, comprising: The satellite hyperspectral imaging payload, one or more processors, and a memory storing program instructions that, when executed by the one or more processors, implement the method described in any one of the first aspects.

[0017] The above-described technical solution of the present invention has the following advantages: This invention provides an imaging deformation correction method, apparatus, and system based on two-way pushbroom velocity estimation. It analyzes the causes of pushbroom imaging deformation of non-cooperative targets in the air and constructs an imaging deformation parameter model. Based on the imaging mechanism of forward and reverse two-way pushbrooms, it analyzes the relative motion difference between the satellite imaging platform and the target under two-way pushbroom imaging and estimates the velocity of the non-cooperative target in the air. Based on this velocity estimation result, combined with the imaging parameters of the satellite payload platform, it solves for the imaging deformation model parameters, determines the spatial resampling frequency of the non-cooperative target in the air, and achieves target morphology reconstruction. This method is a data-driven deformation correction method that does not rely on prior target information and adaptively estimates motion parameters and corrects deformation through the inherent differences in two-way pushbroom data. Attached Figure Description

[0018] The accompanying drawings are provided for illustrative purposes only, and the proportions and quantities of the components in the drawings may not be consistent with the actual product.

[0019] Figure 1 This is a schematic diagram of the pushbroom imaging effect of target satellites at different relative velocities (1 to 4 times). Figure 2 This is a schematic diagram of the relative motion decomposition of a moving target in the bidirectional push-broom mode in an embodiment of the present invention. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] The imaging deformation correction method based on bidirectional push-broom velocity estimation provided in this invention includes the following steps: Step S1: Based on the pushbroom imaging mechanism of the satellite hyperspectral payload, analyze the physical causes of target imaging deformation caused by the relative motion between the satellite platform and the non-cooperative target in the air along the pushbroom direction. That is, the fundamental cause of the deformation is the equivalent spatial resolution of the target along the pushbroom direction. Vertical field of view spatial resolution The ratio of the two is the scaling ratio. .when Deformation occurs at that time; see [reference needed]. Figure 1 Based on this, a target scaling factor model is established to quantitatively describe the deformation.

[0022] Step S2: Load the forward and reverse pushbroom imaging data of the same non-cooperative aerial target acquired by the satellite hyperspectral imaging payload, and simultaneously analyze the imaging parameters transmitted from the satellite platform. Then, use image segmentation to extract the target from these two sets of data, and obtain two key geometric parameters of the target in the pushbroom direction through analysis: the geometric size of the first pixel. Second pixel geometry .

[0023] Step S3: Based on the imaging parameters and pixel geometry obtained in Step S2, conduct an in-depth analysis of the complex relative motion between the satellite platform and the non-cooperative aerial target in both forward and reverse pushbroom modes, and calculate the velocity of the non-cooperative aerial target along the pushbroom direction accordingly. This speed is the core input for subsequent precise deformation correction.

[0024] Step S4: The speed obtained in step S3 Substitute the deformation parameter model established in step S1 to calculate the accurate target scaling factor. Finally, guided by this coefficient, the infrared hyperspectral image data of the mask area where the non-cooperative aerial target is located is spatially resampled along the horizontal push-broom direction, ultimately achieving accurate correction of the target shape.

[0025] This embodiment provides a deformation correction method that does not rely on prior target information and is entirely based on data-driven methods. It adaptively estimates motion parameters and corrects deformation by leveraging the inherent differences in bidirectional push-broom data.

[0026] In one example, in step S1, the constructed target scaling factor model is defined by the following mathematical formula: (1) in: (2) (3) In the formula: The zoom ratio for pushbroom imaging of non-cooperative aerial targets is the direct basis for correction. The equivalent spatial resolution of the target along the push-broom direction depends on the relative motion and the imaging period.

[0027] The spatial resolution of the field of view in the vertical push-broom direction is determined by the system's geometric parameters.

[0028] The relative velocity between the satellite imaging platform and the non-cooperative aerial target along the push-broom direction.

[0029] The imaging period of the satellite hyperspectral imaging payload; The distance between the satellite's hyperspectral imaging payload and non-cooperative targets in the air can be obtained through means such as a spaceborne laser rangefinder. If this is not possible, the satellite's orbital altitude can be used.

[0030] This represents the instantaneous field of view of the satellite's hyperspectral payload.

[0031] This example establishes a precise quantitative relationship model from relative motion, platform parameters to imaging deformation using formulas (1) to (3).

[0032] In one example, in step S2, the geometry of the first pixel is extracted. Second pixel geometry The specific process is as follows: First, the target is separated from the background using image segmentation methods on the loaded forward and backward pushbroom imaging data (usually a multi-channel hyperspectral image cube). In a specific implementation, one or more spectral channels with the highest contrast between the target and the background can be selected, and a threshold segmentation method (such as Otsu's maximum inter-class variance method) can be used for segmentation to generate a binarized target mask image.

[0033] Then, connected component analysis is performed on the generated target mask image to identify and mark pixel regions belonging to non-cooperative aerial targets. Finally, the length (in pixels) of the circumscribed rectangle of this connected component in the push-broom direction is calculated to obtain the results. and .

[0034] In this example, as another optional implementation, the image segmentation method can also employ an edge segmentation method (such as the Canny operator), first extracting the contour of the target, and then obtaining its pixel size in the sweep direction by analyzing the geometric characteristics of the contour.

[0035] This example provides specific methods for automated and robust target information extraction, ensuring the quality of key input data. and The accurate acquisition of this information laid a reliable foundation for subsequent calculations.

[0036] In one example, in step S3, the relative motion between the satellite platform and the non-cooperative aerial target is analyzed (see...). Figure 2 This is a prerequisite for velocity estimation. The specific relative motion relationship is as follows: During forward pushbroom operation, the relative velocities between the satellite platform and the non-cooperative aerial target along the pushbroom direction satisfy the following: (4) During reverse pushbroom operation, the relative velocities between the satellite platform and the non-cooperative airborne target along the pushbroom direction satisfy: (5) in, The scanning speed of the satellite hyperspectral payload platform along the push-broom direction. The velocity of the non-cooperative aerial target along the push-broom direction.

[0037] This example clearly defines the vector relationships between relative velocity, platform velocity, and target velocity in two imaging modes.

[0038] Based on the relative motion relationships suggested in the above examples, the differences in target imaging effects under bidirectional pushbroom scanning are further analyzed. In one example, the pixel size of the target in the image is related to its physical size and equivalent spatial resolution, from which the following relationship can be established: (6) in, It is the projected length of the non-cooperative aerial target in the push-broom direction, which is an unknown quantity.

[0039] Solve the system of equations (6) simultaneously to eliminate the unknowns. Therefore, the formula for directly calculating the velocity of non-cooperative aerial targets can be derived as follows: (7) This example provides insights from observable pixel size differences. and and known platform speed Directly calculate the velocity of non-cooperative aerial targets (unknown targets). Mathematical tools.

[0040] It is worth noting that in this embodiment, formula (7) is a theoretical analysis. Since the speed of the non-cooperative target in the air is unknown, the speed of the non-cooperative target in the air is analyzed by imaging analysis, combined with image features and imaging parameters. In essence, it is the estimated speed of the non-cooperative target in the air obtained by solving the problem. It may have a certain error with the real speed.

[0041] In one example, in step S4, the final target scaling factor needs to be calculated for correction. First, based on the bidirectional push-broom motion analysis and considering the differences in imaging effects, the relative velocity of the target along the push-broom direction is estimated. Then, combined with other known parameters, the following system of equations is used for calculation: (8) Equation (8) gives the value from the estimated velocity of the non-cooperative aerial target to the deformation parameters. The complete computational path transforms the estimated velocity of the non-cooperative aerial target into a deformation scaling parameter that can be directly used for image resampling, thus completing the key transformation from parameter estimation to correction execution.

[0042] It should be noted that the relative velocity in a single push direction of the target can be used during calibration. Specifically, the relative velocity in the forward push direction of the target can be used. Alternatively, the relative velocity in the reverse push-broom direction can be used. See formula (8).

[0043] In one example, in the final stage of step S4, based on the calculated target scaling factor... Spatial resampling along the push-broom direction is performed on the infrared hyperspectral image data within the target mask area.

[0044] In this example, linear interpolation is used for resampling as a preferred method. This method has a moderate computational cost and can effectively smooth out jagged edges while maintaining good spectral characteristics, making it a preferred method for achieving high-precision deformation correction.

[0045] As an alternative, nearest neighbor interpolation can be used. This method is the fastest to compute and can completely preserve the original pixel values, making it suitable for scenarios where processing speed is extremely important and the smoothness of the corrected image is not critical.

[0046] This example provides specific image processing techniques for putting deformation parameters into practice, ensuring the effectiveness and flexibility of deformation correction operations.

[0047] This embodiment also provides an imaging deformation correction device based on bidirectional push-broom velocity estimation, the device comprising: Model building module: used to perform the operation as described in step S1 of Example 1, that is, to build an imaging deformation parameter model.

[0048] Data processing module: Used to perform the operation described in step S2 of Example 1, namely, to complete the imaging parameter analysis and target bidirectional pushbroom data processing, and extract... and .

[0049] Velocity estimation module: Used to perform the operation described in step S3 of Example 1, that is, to complete the bidirectional push-broom relative motion analysis and velocity estimation between the imaging platform and the target, and output the result. .

[0050] Correction execution module: used to perform the operation as described in step S4 of Example 1, that is, to complete the target imaging deformation correction.

[0051] These modules can be implemented through dedicated hardware circuits, programmable logic devices (such as FPGAs), or software programs running on a processor, and can be integrated into satellite ground processing systems or dedicated image processing workstations.

[0052] This device materializes the correction method of the present invention into a device with a clear structure and well-defined functions, clarifying its modular composition, which is beneficial for system integration, testing and application.

[0053] This embodiment also provides an imaging deformation correction system based on bidirectional push-broom velocity estimation, the system comprising: Satellite hyperspectral imaging payload: Used to perform forward and backward pushbroom imaging of non-cooperative targets in the air in orbit to acquire raw hyperspectral data.

[0054] One or more processors: deployed at a ground data processing center.

[0055] Memory: Communicatively connected to the processor, storing computer program instructions.

[0056] When the program instructions are executed by the one or more processors, the system is controlled to implement the correction method as described in any of the above embodiments.

[0057] Any aspects not described in detail in this invention are common knowledge or prior art in the field.

[0058] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that not every embodiment contains only one independent technical solution, and in the absence of conflict between solutions, the various technical features mentioned in each embodiment can be combined in any way to form other implementation methods that can be understood by those skilled in the art.

[0059] Furthermore, without departing from the scope of the present invention, modifications to the technical solutions described in the foregoing embodiments, or equivalent substitutions of some of the technical features, shall not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An imaging deformation correction method based on bidirectional push-broom velocity estimation, characterized in that, Includes the following steps: S1. Based on the pushbroom imaging mechanism of the satellite hyperspectral payload, a target scaling factor model is established to characterize the target imaging deformation caused by the relative motion between the satellite platform and the non-cooperative target in the air along the pushbroom direction. The target scaling factor model is as follows: in, , ; In the formula, Zoom ratio for pushbroom imaging of non-cooperative aerial targets. The equivalent spatial resolution of the target along the push-broom direction, This represents the spatial resolution of the field of view in the vertical push-broom direction. The relative velocity between the satellite imaging platform and the non-cooperative aerial target along the push-broom direction. For the imaging period of the satellite hyperspectral imaging payload, The distance between the satellite's hyperspectral imaging payload and a non-cooperative target in the air. The instantaneous field of view of the satellite's hyperspectral payload; S2. Acquire the forward pushbroom imaging data and backward pushbroom imaging data of the non-cooperative aerial target, as well as the imaging parameters of the satellite platform; extract the first pixel geometric dimension of the non-cooperative aerial target in the pushbroom direction from the forward and backward pushbroom imaging data, respectively. Second pixel geometry ; S3. Based on the imaging parameters of the satellite platform, the geometric dimensions of the first pixel and the second pixel, analyze the relative motion relationship between the satellite platform and the non-cooperative aerial target under forward and reverse pushbroom conditions, and solve for the velocity of the non-cooperative aerial target along the pushbroom direction. The relative motion relationship satisfies: During forward pushbrooming, the relative velocity between the satellite platform and the non-cooperative aerial target along the pushbroom direction. ; During reverse pushbroom operation, the relative velocity between the satellite platform and the non-cooperative aerial target along the pushbroom direction. ; in, The scanning speed of the satellite hyperspectral payload platform along the push-broom direction. The velocity of the non-cooperative aerial target along the push-broom direction; Analysis of the differences in imaging effects of non-cooperative aerial targets under bidirectional pushbroom scanning reveals the following relationship between target pixel sizes: in, The projected length of the non-cooperative aerial target in the push-broom direction; Based on the geometric pixel extraction results of the non-cooperative aerial target under bidirectional pushbroom imaging, and combined with the established geometric pixel relative relationships of the non-cooperative aerial target under bidirectional pushbroom imaging, the velocity of the non-cooperative aerial target along the pushbroom direction is calculated using the following formula. : ； S4, the speed Substituting into the target scaling factor model, the target scaling factor is calculated. Based on the target scaling factor The infrared hyperspectral image data of the mask area where the non-cooperative aerial target is located is spatially resampled along the push-broom direction to correct imaging distortion.

2. The method according to claim 1, characterized in that: In step S2, the non-cooperative aerial target is extracted from the forward and backward pushbroom imaging data using an image segmentation method to generate a target mask image, and the geometric dimensions of the first pixel are obtained through connected component analysis. Second pixel geometry .

3. The method according to claim 2, characterized in that: The image segmentation method is either a threshold segmentation method or an edge segmentation method.

4. The method according to claim 1, characterized in that: The target push-broom imaging scaling ratio is calculated using the following formula. : 。 5. The method according to claim 1 or 4, characterized in that: In step S4, the spatial resampling uses nearest neighbor interpolation or linear interpolation.

6. An imaging deformation correction device based on bidirectional push-broom velocity estimation, used to implement the imaging deformation correction method of claim 1, characterized in that, include: The model building module is used to build an imaging deformation parameter model: based on the pushbroom imaging mechanism of the satellite hyperspectral payload, a target scaling ratio model is established to characterize the target imaging deformation caused by the relative motion between the satellite platform and the non-cooperative target in the air along the pushbroom direction. The data processing module is used for imaging parameter analysis and target two-way pushbroom data processing: acquiring the forward pushbroom imaging data and reverse pushbroom imaging data of the non-cooperative aerial target, as well as the imaging parameters of the satellite platform; Extract the first pixel geometry of the non-cooperative aerial target in the pushbroom direction from the forward and reverse pushbroom imaging data, respectively. Second pixel geometry ; The velocity estimation module is used for target velocity estimation: based on the imaging parameters of the satellite platform, the geometry of the first pixel, and the geometry of the second pixel, it analyzes the relative motion between the satellite platform and the non-cooperative aerial target under forward and reverse pushbroom conditions, and solves for the velocity of the non-cooperative aerial target along the pushbroom direction. ; The correction execution module is used for deformation correction: [The sentence fragment "the speed" appears to be incomplete and lacks context. It's unclear what "the speed" refers to.] Substituting into the target scaling factor model, the target scaling factor is calculated. Based on the target scaling factor The infrared hyperspectral image data of the mask area where the non-cooperative aerial target is located is spatially resampled along the push-broom direction to correct imaging distortion.

7. An imaging deformation correction system based on bidirectional push-broom velocity estimation, characterized in that, include: The satellite hyperspectral imaging payload, one or more processors, and a memory storing program instructions that, when executed by the one or more processors, implement the method as described in any one of claims 1 to 5.

Images