CN115755054A - Low-earth-orbit satellite synthetic aperture radar imaging enhancement method and device - Google Patents

Abstract

The embodiment of the application relates to the field of SAR imaging, and discloses a method and equipment for enhancing low-earth-orbit satellite synthetic aperture radar imaging. The method comprises the following steps: acquiring a first SAR image of a mapping area; segmenting the first SAR image to obtain at least two first sub-images; filling pixels in at least one first sub-image of the at least two first sub-images according to a preset filling rule to obtain at least one second sub-image; and splicing according to at least one second sub-image to obtain a second SAR image of the mapping area. The first SAR image is an SAR image to be optimized, the resolution of each second sub-image is higher than that of the first sub-image corresponding to the second sub-image, and based on the resolution, the resolution of the second SAR image is higher than that of the first SAR image. Therefore, the SAR image can be further optimized by adopting the implementation mode.

Description

Low-earth-orbit satellite synthetic aperture radar imaging enhancement method and device

Technical Field

The embodiment of the application relates to the technical field of image processing, in particular to a Synthetic Aperture Radar (SAR) imaging enhancement optimization method and equipment for a low-orbit satellite.

Background

In the SAR imaging, an antenna is used to radiate electromagnetic waves (e.g., a birdsong (chirp) signal) during the process of SAR movement along a long linear array, and then echo signals of the electromagnetic waves corresponding to different positions are received, and coherent processing is performed on the corresponding echo signals to obtain an SAR image.

Because the SAR is usually carried on an airplane (i.e., airborne) or a satellite (i.e., satellite-borne), and the SAR imaging technology directly processes radar echo signals through an imaging algorithm to obtain an image, it is difficult to obtain a scheme for optimizing the SAR image.

Disclosure of Invention

The embodiment of the application provides a low-earth-orbit satellite synthetic aperture radar imaging enhancement method and equipment, which can further optimize SAR images.

In a first aspect, an embodiment of the present application provides a method for enhancing low-earth-orbit satellite SAR imaging, where the method includes:

acquiring a first SAR image of a mapping area;

segmenting the first SAR image to obtain at least two first sub-images, wherein the width of each first sub-image is smaller than or equal to a preset width threshold value, and the height of each first sub-image is smaller than or equal to a preset height threshold value;

filling pixels in at least one first sub-image of the at least two first sub-images according to a preset filling rule to obtain at least one second sub-image, wherein the resolution of each second sub-image is higher than that of the first sub-image corresponding to the second sub-image;

and splicing according to the at least one second sub-image to obtain a second SAR image of the mapping area.

In some possible embodiments, the filling pixels in at least one of the at least two first sub-images according to a preset filling rule includes:

for each first sub-image to be filled, predicting the pixel value of a pixel to be filled according to the pixel value adjacent to the pixel position to be filled;

and filling the pixels at corresponding positions according to the pixel values of the pixels to be filled.

In some possible embodiments, the filling pixels in at least one of the at least two first sub-images according to a preset filling rule includes:

determining a first sub-image containing a target image in the at least two first sub-images;

and filling pixels in the first sub-image containing the target image according to the preset filling rule.

In some possible embodiments, the method further comprises: and training a filling model to call the filling model to fill pixels in at least one first sub-image of the at least two first sub-images according to a preset filling rule.

In some possible embodiments, the training of the filler model comprises:

acquiring a training sample set, wherein each training sample in the training sample set comprises a first mapping image and a second mapping image of the same target area, and the resolution of the first mapping image is smaller than that of the second mapping image;

and training a neural network by using the training sample set so that the neural network learns the filling rule to obtain the filling model.

In some possible embodiments, the obtaining a training sample set includes:

scanning a to-be-painted area by adopting a first scanning mode to obtain a mapping image of the first area, scanning the to-be-painted area by adopting a second scanning mode to obtain a mapping image of the second area, wherein the resolution of the mapping image of the second area is higher than that of the mapping image of the first area, and the scanning mode is determined according to at least one of the beam width and the Pulse Repetition Frequency (PRF);

determining a region where the first region and the second region overlap as the target region;

and taking the mapping image of the target area in the mapping image of the first area as the first mapping image, and taking the mapping image of the target area in the mapping image of the second area as the second mapping image to obtain a training sample.

In some possible embodiments, the scanning the area to be drawn with the first scanning mode to obtain the mapping image of the first area, and scanning the area to be drawn with the second scanning mode to obtain the mapping image of the second area, includes:

scanning a to-be-painted area by adopting a first wave beam and a first PRF (pulse repetition frequency) to obtain a mapped image of the first area, scanning the to-be-painted area by adopting a second wave beam and a second PRF to obtain a mapped image of the second area, wherein the width of the first wave beam is greater than that of the second wave beam, and the first PRF is lower than that of the second PRF; or,

and scanning the to-be-painted area to be detected in the first scanning direction by adopting a third wave beam to obtain a mapping image of the first area, and scanning the to-be-painted area to be detected in the second scanning direction by adopting the third wave beam to obtain a mapping image of the second area, wherein the first scanning direction is opposite to the second scanning direction.

In some possible embodiments, the scanning the area to be drawn by using the first beam and the first PRF to obtain a mapping image of the first area, and scanning the area to be drawn by using the second beam and the second PRF to obtain a mapping image of the second area includes:

one boundary line of the first area and the second area is coincident; or,

the center point of the first region coincides with the center point of the second region.

In a second aspect, an embodiment of the present application further provides an SAR, where the SAR includes: a transceiver and a processor, wherein the transceiver is connected to the processor,

the transceiver is used for acquiring a first SAR image of a mapping area;

the processor is used for segmenting the first SAR image to obtain at least two first sub-images, wherein the width of each first sub-image is smaller than or equal to a preset width threshold value, and the height of each first sub-image is smaller than or equal to a preset height threshold value;

the processor is further configured to fill pixels in at least one first sub-image of the at least two first sub-images according to a preset filling rule to obtain at least one second sub-image, and a resolution of each second sub-image is higher than a resolution of the first sub-image corresponding to the second sub-image;

the processor is further configured to obtain a second SAR image of the mapping area according to the at least one second sub-image.

In some possible embodiments, the processor is further configured to predict, for each first sub-image to be filled, a pixel value of a pixel to be filled according to a pixel value adjacent to the pixel to be filled; and filling the pixels at corresponding positions according to the pixel values of the pixels to be filled.

In some possible embodiments, the processor is further configured to determine a first sub-image of the at least two first sub-images, which includes the target image; and filling pixels in the first sub-image containing the target image according to the preset filling rule.

In some possible embodiments, the processor is further configured to train a fill model to invoke the fill model to fill pixels in at least one of the at least two first sub-images according to a preset fill rule.

In some possible embodiments, the transceiver is further configured to acquire a set of training samples, each training sample in the set of training samples including a first mapping image and a second mapping image of a same target area, a resolution of the first mapping image being smaller than a resolution of the second mapping image;

the processor is further configured to train a neural network using the training sample set, so that the neural network learns the filling rule to obtain the filling model.

In some possible embodiments, the transceiver is further configured to scan a region to be scanned using a first scanning mode to obtain a mapping image of the first region, scan the region to be scanned using a second scanning mode to obtain a mapping image of the second region, where a resolution of the mapping image of the second region is higher than a resolution of the mapping image of the first region, and the scanning mode is determined according to at least one of a beam width and a pulse repetition frequency PRF;

the processor is further configured to determine an area where the first area and the second area overlap as the target area; and taking the mapping image of the target area in the mapping image of the first area as the first mapping image, and taking the mapping image of the target area in the mapping image of the second area as the second mapping image to obtain a training sample.

In some possible embodiments, the transceiver is further configured to scan a region to be mapped with a first beam and a first PRF to obtain a mapping image of the first region, scan the region to be mapped with a second beam and a second PRF to obtain a mapping image of the second region, where a width of the first beam is greater than a width of the second beam, and the first PRF is lower than the second PRF;

the transceiver is further configured to scan a to-be-painted area in a first scanning direction by using a third beam to obtain a mapping image of the first area, and scan the to-be-painted area in a second scanning direction by using the third beam to obtain a mapping image of the second area, where the first scanning direction is opposite to the second scanning direction.

In some possible embodiments, the scanning the area to be drawn by using the first beam and the first PRF to obtain a mapping image of the first area, and scanning the area to be drawn by using the second beam and the second PRF to obtain a mapping image of the second area includes:

one boundary line of the first area and the second area is coincident; or,

the center point of the first region coincides with the center point of the second region.

In a third aspect, an embodiment of the present application further provides a storage medium, on which computer-executable instructions are stored, and when the computer-executable instructions are executed, the computer is caused to implement the method described in the first aspect or any possible implementation manner of the first aspect.

According to the method and the device for enhancing the imaging of the low-orbit satellite synthetic aperture radar, the pixel filling rule is deployed in advance. Thereafter, a first SAR image of the mapping region is acquired, wherein the first SAR image is a SAR image to be optimized. And then, segmenting the first SAR image to obtain at least two first sub-images, and filling pixels in at least one of the at least two first sub-images according to a filling rule to obtain at least one second sub-image, so that the resolution of each second sub-image is higher than that of the first sub-image corresponding to the second sub-image. And then, splicing at least one second sub-image to obtain a second SAR image of the mapping area, wherein the resolution ratio of the second SAR image is higher than that of the first SAR image. Therefore, by adopting the implementation mode, the SAR image can be further optimized, so that the imaging performance of the SAR can be improved.

Drawings

In order to more clearly describe the technical solutions of the embodiments of the present application, the drawings required to be used in the embodiments are briefly described below, and it is obvious for those skilled in the art to obtain other drawings without creative efforts.

Fig. 1 is a schematic diagram of an exemplary scene of an on-board SAR imaging provided by an embodiment of the present application;

fig. 2 is a flowchart of a method of a SAR image optimization method 100 according to an embodiment of the present application;

FIG. 3A is a schematic diagram of an exemplary scanning scenario for obtaining training samples according to an embodiment of the present disclosure;

FIG. 3A-1 is a schematic illustration of mapping image relationships using the scanning mode illustrated in FIG. 3A;

FIG. 3B is a diagram illustrating a second exemplary scanning scenario for obtaining training samples according to an embodiment of the present disclosure;

FIG. 3B-1 is a schematic drawing of a mapping image relationship using the scanning mode illustrated in FIG. 3B;

FIG. 3C is a diagram illustrating a third exemplary scanning scenario for obtaining training samples according to an embodiment of the present disclosure;

FIG. 3C-1 is a schematic drawing of a mapping image relationship using the scanning mode illustrated in FIG. 3C;

fig. 4 is an exemplary structural schematic diagram of a SAR41 provided in an embodiment of the present application.

Detailed Description

The terminology used in the following examples of the present application is for the purpose of describing alternative embodiments and is not intended to be limiting of the present application. As used in the specification of the present application and the appended claims, the singular forms "a", "an", "the" and "the" are intended to include the plural forms as well. It should also be understood that although the terms first, second, etc. may be used in the following embodiments to describe a class of objects, the objects are not limited to these terms. These terms are used to distinguish between particular objects of that class of objects. For example, the terms first, second, etc. may be used in the following embodiments to describe the SAR image, but the SAR image should not be limited to these terms. The following embodiments may adopt the terms first, second, etc. to describe other class objects in the same way, and are not described herein again.

The following explains implementation scenarios and technical terms related to the embodiments of the present application.

Taking a satellite-borne SAR as an example, refer to a satellite-borne SAR imaging scene diagram illustrated in fig. 1, wherein a satellite transmits an electromagnetic wave to the ground through an antenna, for example, and the electromagnetic wave strikes the ground to generate a reflection signal, and the reflection signal provides imaging data for radar positioning, radar imaging, and the like. The corresponding wave beam of the electromagnetic wave on the ground is a mapping strip, the wave beam width is also the width of the mapping strip, and the image information of the SAR distance direction is represented. As the satellite moves, the swaths on the ground move in respective directions of movement, the direction of satellite movement characterizing the image information of the SAR azimuth direction. And the SAR images according to the image information of the azimuth direction and the image information of the distance direction of the mapping area.

The electromagnetic waves transmitted by the satellite may be called Linear Frequency Modulation (LFM) signals, and may also be called chirp signals. The satellite may transmit a chirp signal at a Pulse frequency repetition (PRF). The LFM signal strikes the ground, producing a reflected signal, which may be referred to as an echo signal.

The embodiment of the application provides a low-orbit satellite SAR imaging enhancement method, which is characterized in that a filling model is trained in advance based on SAR images with different resolutions in the same region, and then pixels are filled in the SAR images to be optimized based on the filling model after the SAR images to be optimized in a surveying and mapping region are obtained, so that the resolution of the SAR images to be optimized is improved, and the optimized SAR images can be obtained.

The technical solutions of the embodiments of the present application are described below with reference to examples.

Referring to fig. 2, fig. 2 shows a SAR image optimization method 100 (hereinafter referred to as method 100) according to an embodiment of the present application, where the method 100 includes the following steps:

step S101, a first SAR image of a mapping area is acquired.

The first SAR image can be obtained by SAR according to a conventional imaging mode, namely, a chirp signal is sent to a corresponding mapping area by adopting a certain PRF, an echo signal of the corresponding chirp signal is received, and then imaging data corresponding to the corresponding echo signal is input to a radar imaging algorithm to obtain the first SAR image. Radar imaging algorithms include, for example: chirp Scaling (CS) algorithm, range Doppler (RD) algorithm, and the like.

In this embodiment, the first SAR image is a SAR image to be optimized.

Step S102, the first SAR image is segmented to obtain at least two first sub-images.

Optionally, in the process of processing the image, the image to be processed may be divided into image blocks according to the maximum image size that can be processed by the image processing function, and then, the processing of the image to be processed is implemented by processing each image block. Similarly, before optimizing the first SAR image, the embodiment of the application may segment the first SAR image to obtain at least two first sub-images, where the width of each first sub-image may be smaller than or equal to a preset width threshold, and the height may be smaller than or equal to a preset height threshold.

Note that the preset width threshold value is a width value of the largest image that can be processed by the image processing function, and the preset height threshold value is a height value of the largest image that can be processed by the image processing function.

Step S103, filling pixels in at least one of the at least two first sub-images according to a preset filling rule to obtain at least one second sub-image.

In some embodiments, pixel filling may be performed on each of the at least two first sub-images to obtain a corresponding second sub-image for each first sub-image. In other embodiments, pixel filling may be performed on a portion of the at least two first sub-images to obtain a second sub-image corresponding to the corresponding portion of the first sub-images, such that the resolution of the second sub-image is higher than the resolution of the corresponding first sub-image. For example, a first sub-image including the target image may be determined from the at least two first sub-images, and then pixels may be filled in the first sub-image including the target image according to a preset filling rule. The target image may be, for example, a predetermined SAR image of the object to be tracked.

Optionally, for each first sub-image to be filled, a pixel value of a pixel to be filled may be predicted according to a pixel value adjacent to a pixel position to be filled, and then the pixel is filled at a corresponding position according to the pixel value of the pixel to be filled.

Optionally, the preset filling rule and the implementation manner of the pixel filling may be implemented by a pre-trained filling model. The filling model may be obtained by training a Neural Network in advance according to a training sample set, and the Neural Network may be, for example, a Convolutional Neural Network (CNN), a Recurrent Neural Network (RNN), or the like.

For example, each training sample in the set of training samples may include a first mapping image and a second mapping image of the same target region, wherein a resolution of the first mapping image is less than a resolution of the second mapping image. Furthermore, the training sample set is used to train the neural network, so that the neural network can learn the filling rules with images of different resolutions in the same target area, thereby obtaining an implementation process of training the neural network to obtain the filling model, which is described in the following embodiments and not described in detail herein.

And step S104, splicing according to at least one second sub-image to obtain a second SAR image of the surveying and mapping area.

In some embodiments, if pixel filling is performed on the at least two first sub-images to obtain a second sub-image corresponding to each first sub-image, the at least one second sub-image may be stitched according to distribution of image content in the first SAR image to obtain a second SAR image. In other embodiments, if pixel filling is performed on a part of the first sub-images in the at least two first sub-images to obtain at least one second sub-image, the at least one second sub-image and the first sub-image without pixel filling may be stitched according to distribution of image content in the first SAR image to obtain a second SAR image.

The number of pixels of the second SAR image is larger than that of the second SAR image, so that the resolution of the second SAR image is larger than that of the second SAR image. Therefore, by adopting the implementation mode, the filling model is obtained by pre-training the SAR images with different resolutions in the same region, and further, after the SAR images in the surveying and mapping region are obtained by processing according to a conventional processing mode, the resolution of the SAR images can be improved by filling pixels based on the filling model, so that the aim of further optimizing the SAR images can be achieved.

According to the above description, the neural network can be trained in advance to obtain the filling model, however, training the neural network requires a large number of training samples, but in the field of SAR imaging technology, only a conventional SAR image of a region can be obtained, and it is difficult to obtain SAR images of different resolutions of a region, so that the difficulty in obtaining the filling model by training is high.

The embodiment of the application provides a method for training a filling model, which comprises the following steps: a set of training samples is acquired, each training sample in the set of training samples comprising a first mapping image and a second mapping image of the same target area, wherein a resolution of the first mapping image may be less than a resolution of the second mapping image. And then, training the neural network by using the training sample set so that the neural network learns filling rules according to a plurality of groups of mapping images with the same region and different resolutions to obtain a filling model.

Wherein, the image resolution in the distance direction is inversely proportional to the pulse width of the LFM signal, and the narrower the pulse width, the higher the PRF of the chirp signal, and the higher the image resolution in the distance direction. The image resolution in azimuth is proportional to the swath width (i.e., beam width), and the larger the swath width, the lower the PRF of the chirp signal and the higher the image resolution in azimuth. Based on this, the SAR may scan the same region to be drawn by using different scanning modes to obtain images of the same region with different resolutions, where the scanning mode is determined according to at least one of the beam width and the PRF.

Optionally, the first scanning mode is used to scan the region to be drawn to obtain a mapping image of the first region, and the second scanning mode is used to scan the region to be drawn to obtain a mapping image of the second region, where the resolution of the mapping image of the second region is higher than that of the mapping image of the first region, for example. The scanning parameters of the first scanning mode and the second scanning mode are different, so that the corresponding sizes of the first area and the second area are different (as shown in fig. 3A, 3B and 3C), further, an area where the first area and the second area overlap is determined as a target area (as shown in fig. 3A-1, 3B-1 and 3C-1), a mapping image of the target area in the mapping image of the first area is used as a first mapping image, and a mapping image of the target area in the mapping image of the second area is used as a second mapping image, so as to obtain a training sample.

In some embodiments, the first beam and the first PRF are used to scan the area to be painted to obtain a first area of the area to be painted, and the second beam and the second PRF are used to scan the area to be painted to obtain a second area of the area to be painted. In this example, the first scanning mode is, for example, using a first beam and a first PRF, the second scanning mode is, for example, using a second beam and a second PRF, the width of the first beam is greater than the width of the second beam (e.g., the beam mapping bandwidth shown in fig. 3A and the beam mapping bandwidth shown in fig. 3B), and the first PRF is lower than the second PRF. According to the imaging principle of the SAR, the resolution of the image obtained by scanning the second beam and the second PRF is greater than the resolution of the image obtained by scanning the first beam and the first PRF.

Illustratively, as shown in fig. 3A, one boundary line of the first beam may be used as a reference, and another boundary line of the beam may be adjusted to obtain the second beam according to the first beam. Thus, as shown in fig. 3A-1, the first region coincides with one boundary line of the second region, and the target region can be acquired with the swath width of the second beam using the coinciding boundary line as a boundary.

For example, as shown in fig. 3B, the second beam may be obtained by narrowing the width of the first beam with the center point of the first beam as the center. Thus, as shown in fig. 3B-1, the center point of the first area coincides with the center point of the second area, and the target area can be acquired by combining the swath width and the azimuth width of the second beam with the center point as the center.

In other embodiments, the third beam is used to scan the drawing area to be measured in the first scanning direction to obtain a surveying and mapping image of the first area, and the third beam is used to scan the drawing area to be measured in the second scanning direction to obtain a surveying and mapping image of the second area, where the first scanning direction is the same as the satellite moving direction, and the first scanning direction is opposite to the second scanning direction.

It should be understood that the second scanning direction is opposite to the first scanning direction, and the first scanning direction is the same direction as the satellite moving direction, accordingly, during the scanning process of the second scanning direction, the speed of each chirp signal pulse moving in the ground direction is lower than the satellite moving speed, and based on this, the resolution of the image obtained by using the second scanning direction is higher than that of the image obtained by using the first scanning direction, that is, the resolution of the mapping image of the second area is higher than that of the mapping image of the first area.

For example, as shown in fig. 3C, the mapping image of the first area obtained by scanning the area to be painted in the first scanning direction is, for example, a radar image obtained in a SAR Stripe Mode (Stripe Mode), and the mapping image of the second area obtained by scanning the area to be painted in the second scanning direction is, for example, a radar image obtained in a SAR bunching Mode. The corresponding beam widths of the mapping image of the first area and the mapping image of the second area are equal, and the azimuth directions are opposite, so that the distance widths of the mapping image of the first area and the mapping image of the second area are the same, and the lengths of the mapping images in the azimuth directions are different in the same time. As shown in fig. 3C-1, the target area may be acquired according to the length of the azimuth direction covered in common in the first scanning direction and the second scanning direction within the same time.

Therefore, by adopting the implementation mode, the SAR images with different resolutions in the same region can be obtained by adopting different scanning modes to scan and image the same region, so that the training data can be flexibly obtained, and the training of the pixel filling model is simple and feasible.

In summary, the low-earth orbit satellite SAR imaging enhancement method provided by the embodiment of the application deploys the pixel filling rule in advance. Thereafter, a first SAR image of the mapping region is acquired, wherein the first SAR image is a SAR image to be optimized. And then, segmenting the first SAR image to obtain at least two first sub-images, and filling pixels in at least one of the at least two first sub-images according to a filling rule to obtain at least one second sub-image, so that the resolution of each second sub-image is higher than that of the first sub-image corresponding to the second sub-image. And then, splicing at least one second sub-image to obtain a second SAR image of the mapping area, wherein the resolution of the second SAR image is higher than that of the first SAR image. Therefore, by adopting the implementation mode, the SAR image can be further optimized, so that the imaging performance of the SAR can be improved.

The above embodiments introduce various embodiments of the SAR image optimization method provided in the embodiments of the present application from the perspective of an action performed by the SAR and the perspective of learning algorithm processing. It should be understood that the embodiments of the present application may implement the above-described functions in hardware or a combination of hardware and computer software corresponding to the above-described process steps. Whether a function is performed as hardware or computer software drives hardware depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.

For example, if the above implementation steps implement the corresponding functions through software modules, the corresponding SAR image optimization apparatus may include a transceiver module and a processing module. The SAR image optimization apparatus may be used to perform some or all of the operations of the method 100 described above.

It is understood that the above division of the modules/units is only a division of logical functions, and in actual implementation, the functions of the above modules may be integrated into a hardware entity, for example, the functions of the processing module may be integrated into a processor, the functions of the transceiver module may be integrated into a transceiver, and programs and instructions for implementing the functions of the above modules may be maintained in a memory. For example, fig. 4 provides a SAR41, the SAR41 comprising may include a processor 411, a transceiver 412, and a memory 413. The transceiver 412 is used for performing transceiving of data and signals in the method 100. The memory 413 may be used for storing programs/codes etc. needed by the processor 411 to execute the method 100. When processor 411 executes code stored in memory 413, SAR41 is caused to perform some or all of the operations of method 100 described above.

E.g., transceiver 412, for acquiring a first SAR image of the mapped region. The processor 411 is configured to segment the first SAR image to obtain at least two first sub-images, where a width of each first sub-image is smaller than or equal to a preset width threshold, and a height of each first sub-image is smaller than or equal to a preset height threshold. The processor 411 is further configured to fill pixels in at least one first sub-image of the at least two first sub-images according to a preset filling rule to obtain at least one second sub-image, where a resolution of each second sub-image is higher than a resolution of the first sub-image corresponding to the second sub-image. The processor 411 is further configured to obtain a second SAR image of the mapping region by stitching according to the at least one second sub-image.

Optionally, the processor 411 is further configured to predict, for each first sub-image to be filled, a pixel value of a pixel to be filled according to a pixel value adjacent to a pixel position to be filled; and filling the pixels at corresponding positions according to the pixel values of the pixels to be filled.

Optionally, the processor 411 is further configured to determine a first sub-image, which includes the target image, of the at least two first sub-images; and filling pixels in the first sub-image containing the target image according to the preset filling rule.

Optionally, the processor 411 is further configured to train a filling model, so as to call the filling model to fill pixels in at least one of the at least two first sub-images according to a preset filling rule.

Optionally, the transceiver 412 is further configured to acquire a training sample set, where each training sample in the training sample set includes a first mapping image and a second mapping image of the same target area, and a resolution of the first mapping image is smaller than a resolution of the second mapping image. The processor 411 is further configured to train a neural network using the training sample set, so that the neural network learns the filling rule to obtain the filling model.

Optionally, the transceiver 412 is further configured to scan the area to be painted by using a first scanning mode to obtain a mapping image of the first area, and scan the area to be painted by using a second scanning mode to obtain a mapping image of the second area, where a resolution of the mapping image of the second area is higher than a resolution of the mapping image of the first area, and the scanning mode is determined according to at least one of a beam width and a pulse repetition frequency PRF. A processor 411, further configured to determine an area where the first area and the second area overlap as the target area; and using the mapping image of the target area in the mapping image of the first area as the first mapping image, and using the mapping image of the target area in the mapping image of the second area as the second mapping image, so as to obtain a training sample.

Optionally, the transceiver 412 is further configured to scan the area to be painted by using a first beam and a first PRF to obtain a mapping image of the first area, scan the area to be painted by using a second beam and a second PRF to obtain a mapping image of the second area, where a width of the first beam is greater than a width of the second beam, and the first PRF is lower than the second PRF. The transceiver 412 is further configured to scan the area to be painted in the first scanning direction by using a third beam to obtain a mapping image of the first area, and scan the area to be painted in the second scanning direction by using the third beam to obtain a mapping image of the second area, where the first scanning direction is opposite to the second scanning direction.

Optionally, the method includes: one boundary line of the first region and the second region is coincident; or the central point of the first area coincides with the central point of the second area.

In specific implementation, corresponding to the aforementioned SAR41, the embodiment of the present application further provides a computer storage medium, where the computer storage medium provided in the SAR41 may store a program, and when the program is executed, part or all of the steps in each embodiment of the method 100 may be implemented. The storage medium may be a magnetic disk, an optical disk, a read-only memory (ROM), a Random Access Memory (RAM), or the like.

One or more of the above modules or units may be implemented in software, hardware or a combination of both. When any of the above modules or units are implemented in software, which is present as computer program instructions and stored in a memory, a processor may be used to execute the program instructions and implement the above method flows. The processor may include, but is not limited to, at least one of: various computing devices that run software, such as a Central Processing Unit (CPU), a microprocessor, a Digital Signal Processor (DSP), a Microcontroller (MCU), or an artificial intelligence processor, may each include one or more cores for executing software instructions to perform operations or processing. The processor may be built in an SoC (system on chip) or an Application Specific Integrated Circuit (ASIC), or may be a separate semiconductor chip. The processor may further include a necessary hardware accelerator such as a Field Programmable Gate Array (FPGA), a PLD (programmable logic device), or a logic circuit for implementing a dedicated logic operation, in addition to a core for executing software instructions to perform an operation or a process.

When the above modules or units are implemented in hardware, the hardware may be any one or any combination of a CPU, a microprocessor, a DSP, an MCU, an artificial intelligence processor, an ASIC, an SoC, an FPGA, a PLD, a dedicated digital circuit, a hardware accelerator, or a discrete device that is not integrated, which may run necessary software or is independent of software to perform the above method flows.

Further, a bus interface may also be included in FIG. 4, which may include any number of interconnected buses and bridges, with one or more processors, represented by a processor, and various circuits of memory, represented by memory, linked together. The bus interface may also link together various other circuits such as peripherals, voltage regulators, power management circuits, and the like, which are well known in the art, and therefore, will not be described any further herein. The bus interface provides an interface. The transceiver provides a means for communicating with various other apparatus over a transmission medium. The processor is responsible for managing the bus architecture and the usual processing, and the memory may store data used by the processor in performing operations.

When the above modules or units are implemented using software, they may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When loaded and executed on a computer, cause the processes or functions described in accordance with the embodiments of the invention to occur, in whole or in part. The computer may be a general purpose computer, a special purpose computer, a network of computers, or other programmable device. The computer instructions may be stored in a computer readable storage medium or transmitted from one computer readable storage medium to another, for example, from one website site, computer, server, or data center to another website site, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital Subscriber Line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device, such as a server, a data center, etc., that includes one or more available media. The usable medium may be a magnetic medium (e.g., floppy Disk, hard Disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid State Disk (SSD)), among others.

It should be understood that, in the various embodiments of the present application, the size of the serial number of each process does not mean the execution sequence, and the execution sequence of each process should be determined by the function and the inherent logic thereof, and should not constitute any limitation to the implementation process of the embodiments.

All parts of the specification are described in a progressive mode, the same and similar parts of all embodiments can be referred to each other, and each embodiment is mainly introduced to be different from other embodiments. In particular, as to the apparatus and system embodiments, since they are substantially similar to the method embodiments, the description is relatively simple and reference may be made to the description of the method embodiments in relevant places.

While alternative embodiments of the present application have been described, additional variations and modifications of these embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, it is intended that the appended claims be interpreted as including the preferred embodiment and all changes and modifications that fall within the scope of the present application.

The above-mentioned embodiments, objects, technical solutions and advantages of the present application are further described in detail, it should be understood that the above-mentioned embodiments are only examples of the present application, and are not intended to limit the scope of the present application, and any modifications, equivalent substitutions, improvements and the like made on the basis of the technical solutions of the present application should be included in the scope of the present invention.

Claims (17)

1. A method for enhancing SAR imaging of a low earth orbit satellite, the method comprising:

acquiring a first SAR image of a mapping area;

segmenting the first SAR image to obtain at least two first sub-images, wherein the width of each first sub-image is smaller than or equal to a preset width threshold value, and the height of each first sub-image is smaller than or equal to a preset height threshold value;

filling pixels in at least one first sub-image of the at least two first sub-images according to a preset filling rule to obtain at least one second sub-image;

and splicing according to the at least one second sub-image to obtain a second SAR image of the mapping area.

2. The method according to claim 1, wherein the filling of pixels in at least one of the at least two first sub-images according to a preset filling rule comprises:

for each first sub-image to be filled, predicting the pixel value of a pixel to be filled according to the pixel value adjacent to the pixel position to be filled;

and filling the pixels at corresponding positions according to the pixel values of the pixels to be filled.

3. The method according to claim 1 or 2, wherein the filling of pixels in at least one of the at least two first sub-images according to a preset filling rule comprises:

determining a first sub-image containing a target image in the at least two first sub-images;

and filling pixels in the first sub-image containing the target image according to the preset filling rule.

4. The method of any of claims 1 to 3, further comprising:

and training a filling model to call the filling model to fill pixels in at least one first sub-image of the at least two first sub-images according to a preset filling rule.

5. The method of claim 4, wherein the training the filler model comprises:

acquiring a training sample set, wherein each training sample in the training sample set comprises a first mapping image and a second mapping image of the same target area, and the resolution of the first mapping image is smaller than that of the second mapping image;

and training a neural network by using the training sample set so that the neural network learns the filling rule to obtain the filling model.

6. The method of claim 5, wherein obtaining the training sample set comprises:

scanning a to-be-painted area by adopting a first scanning mode to obtain a mapping image of the first area, scanning the to-be-painted area by adopting a second scanning mode to obtain a mapping image of the second area, wherein the resolution of the mapping image of the second area is higher than that of the mapping image of the first area, and the scanning mode is determined according to at least one of the beam width and the Pulse Repetition Frequency (PRF);

determining a region where the first region and the second region overlap as the target region;

and taking the mapping image of the target area in the mapping image of the first area as the first mapping image, and taking the mapping image of the target area in the mapping image of the second area as the second mapping image to obtain a training sample.

7. The method as claimed in claim 6, wherein scanning the area to be painted with the first scanning mode to obtain a mapping image of the first area, and scanning the area to be painted with the second scanning mode to obtain a mapping image of the second area comprises:

scanning a to-be-detected drawing area by adopting a first wave beam and a first PRF (pulse repetition frequency) to obtain a drawing image of a first area, scanning the to-be-detected drawing area by adopting a second wave beam and a second PRF to obtain a drawing image of a second area, wherein the width of the first wave beam is greater than that of the second wave beam, and the first PRF is lower than that of the second PRF; or,

and scanning the to-be-painted area to be detected in the first scanning direction by adopting a third wave beam to obtain a mapping image of the first area, and scanning the to-be-painted area to be detected in the second scanning direction by adopting the third wave beam to obtain a mapping image of the second area, wherein the first scanning direction is opposite to the second scanning direction.

8. The method of claim 7, wherein scanning the area to be painted with the first beam and the first PRF to obtain a mapping image of the first area, and scanning the area to be painted with the second beam and the second PRF to obtain a mapping image of the second area comprises:

one boundary line of the first region and the second region is coincident; or,

the center point of the first region coincides with the center point of the second region.

9. A low-orbit satellite Synthetic Aperture Radar (SAR), comprising: a transceiver and a processor, wherein the transceiver is connected to the processor,

the transceiver is used for acquiring a first SAR image of a mapping area;

the processor is used for segmenting the first SAR image to obtain at least two first sub-images, wherein the width of each first sub-image is smaller than or equal to a preset width threshold value, and the height of each first sub-image is smaller than or equal to a preset height threshold value;

the processor is further configured to fill pixels in at least one first sub-image of the at least two first sub-images according to a preset filling rule to obtain at least one second sub-image, and a resolution of each second sub-image is higher than a resolution of the first sub-image corresponding to the second sub-image;

the processor is further configured to obtain a second SAR image of the mapping region by stitching according to the at least one second sub-image.

10. The SAR according to claim 9,

the processor is further configured to predict, for each first sub-image to be filled, a pixel value of a pixel to be filled according to a pixel value adjacent to the pixel to be filled; and filling the pixels at corresponding positions according to the pixel values of the pixels to be filled.

11. The SAR according to claim 9 or 10,

the processor is further configured to determine a first sub-image of the at least two first sub-images that includes the target image; and filling pixels in the first sub-image containing the target image according to the preset filling rule.

12. The SAR of any one of claims 9 to 11, wherein the processor is further configured to train a fill model to invoke the fill model to fill pixels in at least one of the at least two first sub-images according to a preset fill rule.

13. The SAR of claim 12,

the transceiver is further configured to acquire a set of training samples, each training sample in the set of training samples including a first mapping image and a second mapping image of a same target area, a resolution of the first mapping image being smaller than a resolution of the second mapping image;

the processor is further configured to train a neural network using the training sample set, so that the neural network learns the filling rule to obtain the filling model.

14. The SAR according to claim 13,

the transceiver is further configured to scan a region to be drawn by using a first scanning mode to obtain a mapping image of the first region, scan the region to be drawn by using a second scanning mode to obtain a mapping image of the second region, where a resolution of the mapping image of the second region is higher than a resolution of the mapping image of the first region, and the scanning mode is determined according to at least one of a beam width and a Pulse Repetition Frequency (PRF);

the processor is further configured to determine an area where the first area and the second area overlap as the target area; and taking the mapping image of the target area in the mapping image of the first area as the first mapping image, and taking the mapping image of the target area in the mapping image of the second area as the second mapping image to obtain a training sample.

15. The SAR according to claim 14,

the transceiver is further configured to scan a to-be-painted area by using a first beam and a first PRF to obtain a mapping image of a first area, scan the to-be-painted area by using a second beam and a second PRF to obtain a mapping image of a second area, where a width of the first beam is greater than a width of the second beam, and the first PRF is lower than the second PRF;

the transceiver is further configured to scan a to-be-painted area in a first scanning direction by using a third beam to obtain a mapping image of the first area, and scan the to-be-painted area in a second scanning direction by using the third beam to obtain a mapping image of the second area, where the first scanning direction is opposite to the second scanning direction.

16. The SAR of claim 15,

one boundary line of the first area and the second area is coincident; or,

the center point of the first region coincides with the center point of the second region.

17. A storage medium having stored thereon computer-executable instructions that, when executed, cause a computer to implement the method of any one of claims 1 to 8.

Images