Radar detection method, device and equipment based on multi-directional gradient enhancement detection

Abstract

This invention provides a radar detection method, apparatus, and device based on multi-directional gradient enhancement detection, which can be applied to the field of radar detection technology. The radar detection method based on multi-directional gradient enhancement detection includes: responding to receiving an electromagnetic echo signal reflected by a target region to be detected, performing radar imaging processing on the electromagnetic echo signal to obtain a detection region image of the target region; performing registration processing on the detection region image and a reference region image corresponding to the target region to obtain a phase interference region image of the target region; performing gradient processing on multiple pixels in the phase interference region image according to multiple preset gradient directions based on a preset neighborhood range to obtain a multi-directional gradient region feature image; and performing segmentation processing on the multi-directional gradient region feature image based on a first gradient threshold to obtain a target detection result corresponding to the target region.

Description

Technical Field

[0001] This invention relates to the field of radar detection technology, and more specifically to a radar detection method, apparatus, and equipment based on multi-directional gradient enhancement detection. Background Technology

[0002] Real-time detection and removal of foreign object debris (FOD) on a site is a crucial aspect of ensuring the safety of transportation and production operations. Since manual, periodic inspections of sites such as airport runways for FOD are inefficient, and small FOD fragments are difficult to observe in harsh environments, radar is typically used to detect FOD on the site.

[0003] In the process of realizing the above-mentioned inventive concept, it was found through research that in the process of using radar to detect tiny foreign object debris in related technologies, due to a large amount of external noise, it is difficult to accurately detect tiny foreign object debris. At the same time, due to the complexity and variability of the environment, devices such as metal railings in the scene are easily misjudged as foreign object debris, thus leading to the technical problem of false alarms. Summary of the Invention

[0004] In view of the above problems, the present invention provides a radar detection method, apparatus and equipment based on multi-directional gradient enhancement detection.

[0005] According to a first aspect of the present invention, a radar detection method based on multi-directional gradient enhancement detection is provided, comprising: responding to receiving an electromagnetic echo signal reflected by a target region to be detected, performing radar imaging processing on the electromagnetic echo signal to obtain a detection region image of the target region; performing registration processing on the detection region image and a reference region image corresponding to the target region to obtain a phase interference region image of the target region; performing gradient processing on multiple pixels in the phase interference region image according to multiple preset gradient directions based on a preset neighborhood range to obtain a multi-directional gradient region feature image, wherein the multi-directional gradient region feature image represents the image obtained by fusing with image features corresponding to each preset gradient direction; and performing segmentation processing on the multi-directional gradient region feature image based on a first gradient threshold to obtain a target detection result corresponding to the target region.

[0006] According to an embodiment of the present invention, a phase interference region image of the target region is obtained by registering a detection region image and a reference region image corresponding to the target region. This includes: determining a detection feature window and a reference feature window from the detection region image and the reference region image respectively, based on feature segmentation rules; constructing a feature cross-correlation function by processing the detection coordinate information and detection amplitude information of multiple detection feature pixels within the detection feature window with the reference coordinate information and reference amplitude information of multiple reference feature pixels within the reference feature window; and iteratively adjusting the detection region image according to the feature cross-correlation function to obtain the phase interference region image.

[0007] According to an embodiment of the present invention, the detection region image is iteratively adjusted according to the feature cross-correlation function to obtain a phase interference region image, including: for the i-th iteration round, where i ≥ 2 and i is a positive integer; adjusting the adjusted detection image of the i-th iteration round using the i-1th first offset information determined from the i-1th feature cross-correlation function to obtain the adjusted detection image of the i-th iteration round, wherein the first offset information is characterized as the position information corresponding to the feature cross-correlation peak in the feature cross-correlation function; determining the i-th detection feature window and the i-th reference feature window from the adjusted detection image and the reference region image of the i-th iteration round, respectively, based on feature partitioning rules; and adjusting the i-th detection region image by adjusting the i-th feature cross-correlation function. The detection coordinates and amplitude information of multiple detection feature pixels within the feature window are processed with the reference coordinates and amplitude information of multiple reference feature pixels within the i-th reference feature window to construct the i-th feature cross-correlation function. The i-th first offset information corresponding to the peak value of the i-th feature cross-correlation function is determined from the i-th feature cross-correlation function. If the i-th first offset information is greater than or equal to a predetermined offset threshold, the above operation is repeated until the i-th first offset information is less than the predetermined offset threshold. If the i-th first offset information is less than the predetermined offset threshold, the i-th first offset information is used to adjust the adjustment detection image of the i-th iteration round to obtain the phase interference region image.

[0008] According to an embodiment of the present invention, a multi-directional gradient region feature image is obtained by performing gradient processing on multiple pixels in a phase interference region image based on a preset neighborhood range and multiple preset gradient directions. The process includes: dividing the phase interference region image based on a preset neighborhood range and multiple preset gradient directions to obtain multiple sub-band regions corresponding to each pixel under each preset gradient direction; performing weighted processing on the multiple sub-band regions corresponding to each preset gradient direction to obtain a unidirectional gradient region feature image corresponding to each preset gradient direction; and performing cumulative multiplication processing on the multiple unidirectional gradient region feature images to obtain a multi-directional gradient region feature image.

[0009] According to an embodiment of the present invention, weighted processing is performed on multiple sub-band regions corresponding to each preset gradient direction to obtain a unidirectional gradient region feature image corresponding to each preset gradient direction, including: for the p-th pixel, where p≥1, p is a positive integer; enhancing the amplitude information in the multiple sub-band regions corresponding to each preset gradient direction to obtain a detection enhancement gradient value corresponding to each preset gradient direction; calculating the standard deviation of the amplitude information in the multiple sub-band regions to obtain a detection weighting operator corresponding to each preset gradient direction; obtaining a target detection gradient value corresponding to each preset gradient direction based on the multiple detection enhancement gradient values ​​and the multiple detection weighting operators; and constructing a unidirectional gradient region feature image corresponding to each preset gradient direction based on the target detection gradient values ​​of P pixels corresponding to each preset gradient direction, where P≥p, p is a positive integer.

[0010] According to an embodiment of the present invention, the amplitude information in multiple sub-band regions corresponding to each preset gradient direction is enhanced to obtain a detection enhancement gradient value corresponding to each preset gradient direction. This includes: averaging the amplitude information of each pixel in the multiple sub-band regions for any given preset gradient direction to obtain the mean amplitude information of the multiple sub-band regions; calculating the difference based on the mean amplitude information of the multiple sub-band regions according to sub-band region rules to obtain a first mean amplitude difference and a second mean amplitude difference, wherein the first mean amplitude difference is calculated from the mean amplitude information of a portion of the multiple sub-band regions, and the second mean amplitude difference is calculated from the mean amplitude information of another portion of the sub-band regions; and taking the minimum value between the first mean amplitude difference and the second mean amplitude difference as the detection enhancement gradient value corresponding to any given preset gradient direction.

[0011] According to an embodiment of the present invention, standard deviation calculation is performed on amplitude information in multiple sub-band regions to obtain a detection weighting operator corresponding to each preset gradient direction. This includes: for any preset gradient direction, based on sub-band region rules, standard deviation calculation is performed on the amplitude information of each pixel in multiple sub-band regions to obtain a first amplitude standard deviation and a second amplitude standard deviation, wherein the first amplitude standard deviation is calculated from the amplitude information of each pixel in a portion of the sub-band regions, and the second amplitude standard deviation is calculated from the amplitude information of each pixel in another portion of the sub-band regions; based on the first amplitude standard deviation and the second amplitude standard deviation, a detection weighting operator corresponding to any preset gradient direction is obtained.

[0012] According to an embodiment of the present invention, the multi-directional gradient region feature image is divided based on a first gradient threshold to obtain a target detection result corresponding to the target region, including: dividing the multi-directional gradient region feature image based on the first gradient threshold to obtain target detection gradient points greater than the first gradient threshold; and obtaining the target detection result corresponding to the target region based on the target detection gradient points and the detection coordinate information of the target detection gradient points.

[0013] A second aspect of the present invention provides a radar detection device based on multi-directional gradient enhancement detection, comprising: an imaging module, configured to perform radar imaging processing on the electromagnetic echo signal reflected by a target region to be detected in response to receiving such an electromagnetic echo signal, thereby obtaining a detection region image of the target region; a registration module, configured to perform registration processing on the detection region image and a reference region image corresponding to the target region, thereby obtaining a phase interference region image of the target region; a gradient processing module, configured to perform gradient processing on multiple pixels in the phase interference region image according to multiple preset gradient directions based on a preset neighborhood range, thereby obtaining a multi-directional gradient region feature image, wherein the multi-directional gradient region feature image represents the image obtained after fusing with image features corresponding to each preset gradient direction; and a segmentation module, configured to perform segmentation processing on the multi-directional gradient region feature image based on a first gradient threshold, thereby obtaining a target detection result corresponding to the target region.

[0014] A third aspect of the present invention provides an electronic device comprising: one or more processors; and a memory for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors perform the method described above.

[0015] A fourth aspect of the present invention also provides a computer-readable storage medium having executable instructions stored thereon, which, when executed by a processor, cause the processor to perform the methods described above.

[0016] A fifth aspect of the invention also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method.

[0017] According to the radar detection method, apparatus, and device based on multi-directional gradient enhancement detection of the present invention, when it is necessary to confirm whether there are tiny foreign object fragments in the target area to be detected, a reference area image of a reference area corresponding to the target area to be detected that does not contain any tiny foreign object fragments is acquired in advance. Then, a radar signal is sent to the target area to be detected. In response to the electromagnetic echo signal reflected back from the target area to be detected, the electromagnetic echo data of each channel in the electromagnetic echo signal is superimposed and jointly imaged to obtain a detection area image of the target area. Then, the detection area image and the reference area image are registered to obtain a phase interference area image of the target area. This realizes the adjustment and correction of the positional shift of image features in the detection area image caused by environmental noise, thereby improving the accuracy of image features and the robustness of identifying and correcting dynamic noise, so as to more accurately identify whether there are tiny foreign object fragments in the image and reduce the probability of misjudgment.

[0018] According to an embodiment of the present invention, after performing gradient processing on multiple pixels in the adjusted phase interference region image based on a preset neighborhood range and multiple preset gradient directions, a multi-directional gradient region feature image containing unidirectional gradient features corresponding to each preset gradient direction is obtained. Based on a first gradient threshold, the presence of tiny foreign object fragments and their location information are determined from the multi-directional gradient region feature image. This achieves feature extraction and fusion from multiple dimensions and angles, fully exploring the multi-directional gradient features of tiny foreign objects in spatial structure. The extracted multi-directional gradient features generate corresponding multi-directional gradient region feature images, and the images are confirmed by combining the first gradient threshold. This further enhances the signal-to-noise ratio of tiny targets, so that even in complex environments, accurate target detection results can be obtained through multi-dimensional feature analysis, improving detection accuracy, reducing the probability of missed detections and false alarms, improving detection efficiency, and reducing labor costs, making it suitable for widespread application in industrial and transportation production. Attached Figure Description

[0019] The above-described features, other objects, and advantages of the present invention will become clearer from the following description of embodiments of the invention with reference to the accompanying drawings, in which:

[0020] Figure 1 An application scenario diagram of the radar detection method based on multi-directional gradient enhancement detection according to an embodiment of the present invention is shown;

[0021] Figure 2 A flowchart of a radar detection method based on multi-directional gradient enhancement detection according to an embodiment of the present invention is shown;

[0022] Figure 3A flowchart illustrating the process of obtaining a phase interference region image of a target region according to an embodiment of the present invention is shown;

[0023] Figure 4 A flowchart illustrating the process of obtaining a multi-directional gradient region feature image according to an embodiment of the present invention is shown;

[0024] Figure 5 A schematic diagram of multiple sub-band regions with a preset gradient direction of 0° is shown according to an embodiment of the present invention;

[0025] Figure 6 A schematic diagram of multiple sub-band regions with a preset gradient direction of 45° is shown according to an embodiment of the present invention;

[0026] Figure 7 A schematic diagram of multiple sub-band regions with a preset gradient direction of 90° is shown according to an embodiment of the present invention;

[0027] Figure 8 A schematic diagram of multiple sub-band regions with a preset gradient direction of 135° is shown according to an embodiment of the present invention;

[0028] Figure 9 A schematic diagram of a multi-directional gradient region feature image according to an embodiment of the present invention is shown;

[0029] Figure 10 A structural block diagram of a radar detection device based on multi-directional gradient enhancement detection according to an embodiment of the present invention is shown;

[0030] Figure 11 A block diagram of an electronic device based on a radar detection method using multi-directional gradient enhancement detection according to an embodiment of the present invention is shown. Detailed Implementation

[0031] Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of the invention. In the following detailed description, numerous specific details are set forth to provide a thorough understanding of the embodiments of the invention for ease of explanation. However, it will be apparent that one or more embodiments may be practiced without these specific details. Furthermore, descriptions of well-known structures and techniques are omitted in the following description to avoid unnecessarily obscuring the concept of the invention.

[0032] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the invention. The terms “comprising,” “including,” etc., as used herein indicate the presence of the stated features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0033] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein are to be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.

[0034] When using expressions such as "at least one of A, B and C", they should generally be interpreted in accordance with the meaning that is commonly understood by those skilled in the art (e.g., "a system having at least one of A, B and C" should include, but is not limited to, a system having A alone, a system having B alone, a system having C alone, a system having A and B, a system having A and C, a system having B and C, and / or a system having A, B and C, etc.).

[0035] In the technical solution of this invention, the user information (including but not limited to user personal information, user image information, user device information, such as location information) and data (including but not limited to data used for analysis, stored data, and displayed data) involved are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, storage, use, processing, transmission, provision, invention, and application of related data all comply with relevant laws, regulations, and standards, take necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation entry points for users to choose to authorize or refuse.

[0036] Real-time detection and removal of foreign object debris on the site is a crucial aspect of ensuring the safety of transportation and production operations. Typically, small foreign object debris on runways can be detected through regular manual patrols; however, this method is inefficient and unreliable. Manual patrols are time-consuming, have a low refresh rate, and are affected by visual observation in foggy or other weather conditions. Furthermore, small metal objects, such as screws, which pose significant safety hazards to aircraft, are difficult to detect in a timely manner.

[0037] Therefore, related technologies typically employ radar to detect foreign object debris on the field. Currently, radar-based systems for detecting small foreign object debris are mainly edge-lamp detection systems. These systems install detectors in an "edge-lamp" configuration on both sides of the runway. However, in the process of using radar to detect small foreign object debris, significant ambient noise makes accurate detection difficult. Furthermore, the complex and variable environment can cause objects such as metal railings to be mistakenly identified as foreign object debris, leading to false alarms.

[0038] In view of this, embodiments of the present invention provide a radar detection method based on multi-directional gradient enhancement detection, comprising: responding to receiving an electromagnetic echo signal reflected by a target region to be detected, performing radar imaging processing on the electromagnetic echo signal to obtain a detection region image of the target region; performing registration processing on the detection region image and a reference region image corresponding to the target region to obtain a phase interference region image of the target region; performing gradient processing on multiple pixels in the phase interference region image according to multiple preset gradient directions based on a preset neighborhood range to obtain a multi-directional gradient region feature image, wherein the multi-directional gradient region feature image represents the image obtained after fusing with image features corresponding to each preset gradient direction; and performing segmentation processing on the multi-directional gradient region feature image based on a first gradient threshold to obtain a target detection result corresponding to the target region.

[0039] Figure 1 An application scenario diagram of the radar detection method based on multi-directional gradient enhancement detection according to an embodiment of the present invention is shown.

[0040] like Figure 1 As shown, the application scenario according to this embodiment may include multiple wireless devices 101, mobile devices 102, a transportation site 103, foreign object debris 104, and a processing device 105. The multiple wireless devices 101 are used to transmit radar signals to the transportation site 103.

[0041] Users can use multiple wireless devices 101 to interact with the processing device 105 via radar signals to obtain corresponding signal information, etc. The processing device 105 can be a computing device that processes the signal information.

[0042] The processing device 105 can analyze and process data such as signal information emitted by the wireless device 101, and feed back the processing results.

[0043] The transport site 103 and foreign object debris 104 can reflect signals to multiple wireless devices 101.

[0044] It should be noted that the radar detection method based on multi-directional gradient enhancement detection provided in the embodiments of the present invention can generally be executed by multiple wireless devices 101. Correspondingly, the radar detection device based on multi-directional gradient enhancement detection provided in the embodiments of the present invention can generally be disposed in multiple wireless devices 101. The radar detection method based on multi-directional gradient enhancement detection provided in the embodiments of the present invention can also be executed by a processing device 105 or a cluster of processing devices that are different from the multiple wireless devices 101. Correspondingly, the radar detection device based on multi-directional gradient enhancement detection provided in the embodiments of the present invention can also be disposed in a processing device 105 or a cluster of processing devices that are different from the multiple wireless devices 101.

[0045] It should be understood that Figure 1 The number of wireless and processing devices shown is merely illustrative. Any number of wireless and processing devices can be used depending on implementation needs.

[0046] The following will be based on Figure 1 The described scene, through Figures 2-8 The radar detection method based on multi-directional gradient enhancement detection according to embodiments of the present invention will be described in detail.

[0047] Figure 2 A flowchart of a radar detection method based on multi-directional gradient enhancement detection according to an embodiment of the present invention is shown.

[0048] like Figure 2 As shown, the radar detection method based on multi-directional gradient enhancement detection in this embodiment includes operations S210~S240.

[0049] In operation S210, in response to receiving the electromagnetic echo signal reflected by the target area to be detected, radar imaging processing is performed on the electromagnetic echo signal to obtain a detection area image of the target area.

[0050] According to an embodiment of the present invention, the target area to be detected can be characterized as an area containing tiny foreign object fragments.

[0051] According to an embodiment of the present invention, radar signals are transmitted to the target area by transmitting antennas of radars positioned on both sides of the target area to be detected. The target area to be detected can generate and reflect electromagnetic echo signals back to the receiving antenna of the radar. When the receiving antenna of the radar receives the electromagnetic echo signals, the electromagnetic echo signals can be sent to a processing device, and then radar imaging processing is performed on the electromagnetic echo signals so that a detection area image corresponding to the target area can be generated based on the electromagnetic echo signals.

[0052] According to an embodiment of the present invention, the electromagnetic echo signal reflected back from the target area to be detected may include multi-channel electromagnetic echo data. The processing device can perform superposition and common imaging processing of the electromagnetic echo data of each channel to obtain the detection area image. The detection area image can be obtained according to formula (1), as shown below.

[0053] (1);

[0054] Where I1(x,y) can be represented as the detection region image, It can be characterized as the electromagnetic echo signal reflected by the target area to be detected, and t can be characterized as the sampling time. The time when the radar's receiving antenna receives the electromagnetic echo signal can be represented by , and m can be represented by the signal received by the m-th receiving antenna. It can be characterized as wavelength. It can be characterized as the first The azimuth position of the m-th receiving antenna at time m. It can be characterized as the first The range position of the m-th receiving antenna at time m, x can be represented as azimuth information, y can be represented as range information, and j can be represented as the imaginary unit.

[0055] In operation S220, the detection region image and the reference region image corresponding to the target region are registered to obtain the phase interference region image of the target region.

[0056] According to an embodiment of the present invention, before acquiring the detection region image of the target region to be detected, it is necessary to first acquire the region image of the reference region corresponding to the target region that is free of any tiny foreign object debris, i.e., the reference region image.

[0057] According to an embodiment of the present invention, a reference area image can be generated by transmitting radar signals to a reference area free of any tiny foreign object debris using radar transmitting antennas positioned on both sides of the target area to be detected. The reference area generates electromagnetic echo signals and reflects these signals back to the radar receiving antenna. Then, radar imaging processing is performed on the electromagnetic echo signals to generate a reference area image based on the electromagnetic echo signals.

[0058] According to an embodiment of the present invention, the electromagnetic echo signal reflected back from a region without any tiny foreign object debris may include multi-channel electromagnetic echo data. After superimposing the electromagnetic echo data of each channel for joint imaging, a reference region image can be obtained. The reference region image can be obtained according to formula (2), as shown below.

[0059] (2);

[0060] Where I0(x, y) can be represented as the reference region image, It can be characterized as an electromagnetic echo signal reflected by a reference region that contains no tiny foreign object debris.

[0061] According to embodiments of the present invention, by performing registration processing on the detection area image and the reference area image, the image shift caused by environmental noise can be reduced, thereby improving the recognition accuracy.

[0062] In operation S230, based on a preset neighborhood range, gradient processing is performed on multiple pixels in the phase interference region image according to multiple preset gradient directions to obtain a multi-directional gradient region feature image.

[0063] According to an embodiment of the present invention, the multi-directional gradient region feature image representation is an image obtained by fusing the image features corresponding to each preset gradient direction.

[0064] According to an embodiment of the present invention, the multi-directional gradient region feature image may include image features corresponding to each preset gradient direction. Thus, by fusing the image features corresponding to each preset gradient direction, the features of tiny foreign object fragments can be represented from multiple levels in the spatial dimension.

[0065] In operation S240, based on the first gradient threshold, the feature image of the multi-directional gradient region is divided to obtain the target detection result corresponding to the target region.

[0066] According to an embodiment of the present invention, the target detection result may include whether there are tiny foreign object fragments in the target area, and the location information of the tiny foreign object fragments.

[0067] According to an embodiment of the present invention, when it is necessary to confirm whether there are tiny foreign object fragments in the target area to be detected, a reference area image of a reference area corresponding to the target area to be detected that does not contain any tiny foreign object fragments is acquired in advance. Then, a radar signal is sent to the target area to be detected. In response to the electromagnetic echo signal reflected back from the target area to be detected, the electromagnetic echo data of each channel in the electromagnetic echo signal is superimposed and jointly imaged to obtain a detection area image of the target area. Then, the detection area image and the reference area image are registered to obtain a phase interference area image of the target area. This realizes the adjustment and correction of the positional shift of image features in the detection area image caused by environmental noise, thereby improving the accuracy of image features and the robustness of identifying and correcting dynamic noise, so as to more accurately identify whether there are tiny foreign object fragments in the image and reduce the probability of misjudgment.

[0068] According to an embodiment of the present invention, after performing gradient processing on multiple pixels in the adjusted phase interference region image based on a preset neighborhood range and multiple preset gradient directions, a multi-directional gradient region feature image containing unidirectional gradient features corresponding to each preset gradient direction is obtained. Based on a first gradient threshold, the presence of tiny foreign object fragments and their location information are determined from the multi-directional gradient region feature image. This achieves feature extraction and fusion from multiple dimensions and angles, fully exploring the multi-directional gradient features of tiny foreign objects in spatial structure. The extracted multi-directional gradient features generate corresponding multi-directional gradient region feature images, and the images are confirmed by combining the first gradient threshold. This further enhances the signal-to-noise ratio of tiny targets, so that even in complex environments, accurate target detection results can be obtained through multi-dimensional feature analysis, improving detection accuracy, reducing the probability of missed detections and false alarms, improving detection efficiency, and reducing labor costs, making it suitable for widespread application in industrial and transportation production.

[0069] According to an embodiment of the present invention, a method for registering a detection region image and a reference region image corresponding to the target region to obtain a phase interference region image of the target region includes the following operations.

[0070] Figure 3 A flowchart illustrating the process of obtaining a phase interference region image of a target region according to an embodiment of the present invention is shown.

[0071] like Figure 3 As shown, obtaining the phase interference region image of the target region in this embodiment includes operations S310 to S330.

[0072] In operation S310, based on feature segmentation rules, detection feature windows and reference feature windows are determined from the detection region image and the reference region image, respectively.

[0073] According to an embodiment of the present invention, the feature segmentation rule can be characterized as segmentation based on the concentrated region of feature pixels in the image.

[0074] According to an embodiment of the present invention, the detection region image is divided based on feature segmentation rules to obtain a region containing a relatively concentrated set of feature pixels, i.e., a detection feature window. At the same time, the reference region image is divided based on feature segmentation rules to obtain a region containing a relatively concentrated set of feature pixels, i.e., a reference feature window.

[0075] In operation S320, a feature cross-correlation function is constructed by processing the detection coordinate information and detection amplitude information of multiple detection feature pixels in the detection feature window and the reference coordinate information and reference amplitude information of multiple reference feature pixels in the reference feature window.

[0076] According to embodiments of the present invention, the feature cross-correlation function is obtained by cross-correlation calculation based on the detection coordinate information and detection amplitude information of each feature pixel in the detection feature window and the reference coordinate information and reference amplitude information of each feature pixel in the reference feature window. Based on the feature cross-correlation function, pixels with offsets in the detection region image can be adjusted.

[0077] In operation S330, the detection region image is iteratively adjusted according to the feature cross-correlation function to obtain the phase interference region image.

[0078] According to an embodiment of the present invention, the method for iteratively adjusting the detection region image based on the feature cross-correlation function to obtain the phase interference region image includes the following operations.

[0079] According to an embodiment of the present invention, for the i-th iteration, where i ≥ 2 and i is a positive integer.

[0080] According to an embodiment of the present invention, the adjustment detection image of the (i-1)th iteration round is adjusted using the first offset information determined from the (i-1)th feature cross-correlation function to obtain the adjustment detection image of the ith iteration round.

[0081] According to an embodiment of the present invention, the first offset information is characterized as the position information corresponding to the characteristic cross-correlation peak in the characteristic cross-correlation function.

[0082] According to an embodiment of the present invention, the first offset information includes azimuth offset information and range offset information. The detection area image and the reference area image are registered multiple times with different precisions, thereby improving the accuracy of feature pixels in the detection area image.

[0083] For example, after constructing the first feature cross-correlation function for the detection region image and the reference region image, the feature cross-correlation peak and the position information corresponding to the feature cross-correlation peak are determined from the first feature cross-correlation function. Then, the first offset information is used to adjust each feature pixel, thereby obtaining the adjusted detection image in the second iteration round.

[0084] According to an embodiment of the present invention, based on feature segmentation rules, the i-th detection feature window and the i-th reference feature window are determined from the adjusted detection image and the reference region image in the i-th iteration, respectively.

[0085] According to an embodiment of the present invention, in a more precise new round of iterative adjustment, it is necessary to re-divide the feature windows of the adjustment detection image and the reference region image in this round of iteration according to the feature segmentation principle.

[0086] For example, in the first iteration, the feature windows of the detection region image and the reference region image are divided and adjusted with a precision of 10. In the second iteration, the adjusted detection image and the unchanging reference region image in the second iteration are divided and adjusted with a precision of 1 to obtain the second detection feature window and the second reference feature window.

[0087] According to an embodiment of the present invention, the i-th feature cross-correlation function is constructed by processing the detection coordinate information and detection amplitude information of multiple detection feature pixels in the i-th detection feature window and the reference coordinate information and reference amplitude information of multiple reference feature pixels in the i-th reference feature window.

[0088] According to an embodiment of the present invention, in a more precise new round of iterative adjustment, the same feature cross-correlation function as described above is constructed for each feature pixel in the i-th detection feature window and each feature pixel in the i-th reference feature window, thereby obtaining the i-th feature cross-correlation function in this round of iteration.

[0089] For example, in the second iteration, a second feature cross-correlation function is constructed based on the detection coordinates and detection amplitude information of each feature pixel in the second detection feature window and the reference coordinates and reference amplitude information of each feature pixel in the second reference feature window.

[0090] According to an embodiment of the present invention, the i-th first offset information corresponding to the peak value of the i-th feature cross-correlation function is determined from the i-th feature cross-correlation function.

[0091] For example, in the second iteration, the second characteristic cross-correlation peak and the second first offset information corresponding to the peak are determined from the second characteristic cross-correlation function.

[0092] According to an embodiment of the present invention, if the i-th first offset information is greater than or equal to a predetermined offset threshold, the above operation is repeated until the i-th first offset information is less than the predetermined offset threshold.

[0093] According to an embodiment of the present invention, when the i-th first offset information is less than a predetermined offset threshold, the i-th first offset information is used to adjust the adjustment detection image of the i-th iteration to obtain a phase interference region image.

[0094] According to an embodiment of the present invention, by using a preset target precision for image adjustment, if the adjustment precision of the detection area image is less than a predetermined offset threshold, iterative adjustment can be performed to obtain the adjusted target detection image.

[0095] According to embodiments of the present invention, in addition to setting a predetermined offset threshold, a preset number of iterations can also be set to adjust the detection region image by the preset number of iterations, so as to make the detection region image more accurate and obtain the adjusted target detection image.

[0096] According to an embodiment of the present invention, when an adjusted target detection image is obtained, a phase interference region image can be obtained based on the adjusted target detection image and the adjusted reference region image, i.e., the original reference region image.

[0097] According to an embodiment of the present invention, the phase interference region image can be as shown in formula (3).

[0098] (3);

[0099] Wherein, P(x,y) can be characterized as the phase interference region image, This can be characterized as phase calculation. It can be represented as a reference region image after multiple rounds of iterative registration. It can be characterized as a target detection image after multiple rounds of iterative registration. It can be characterized as the conjugate of the target detection image after multiple rounds of iterative registration.

[0100] According to an embodiment of the present invention, based on feature segmentation rules, the i-th detection feature window and the i-th reference feature window are determined from the detection region image and the reference region image of the i-th iteration, respectively. Then, the i-th feature cross-correlation function is constructed according to the coordinate and amplitude information of each feature pixel in the i-th detection feature window and the i-th reference feature window. The i-th first offset information is determined according to the feature cross-correlation peak value in the i-th feature cross-correlation function. The i-th first offset information is used to adjust the reference region image, and it is determined whether the predetermined offset threshold for stopping the iterative adjustment is met. This realizes multi-round iterative registration adjustment between the detection region image and the reference region image, and round-by-round registration from low precision to high precision. This maximizes the coherence between the adjusted target detection image and the reference region image, reduces the impact of imaging offset caused by environmental noise, and improves the accuracy and efficiency of detection and recognition.

[0101] According to embodiments of the present invention, a method for obtaining a multi-directional gradient region feature image by performing gradient processing on multiple pixels in a phase interference region image based on an angle focusing criterion, a preset neighborhood range, and multiple preset gradient directions includes the following operations.

[0102] Figure 4 A flowchart illustrating the process of obtaining a multi-directional gradient region feature image according to an embodiment of the present invention is shown.

[0103] like Figure 4 As shown, obtaining the phase interference region image of the target region in this embodiment includes operations S410 to S430.

[0104] In operation S410, the phase interference region image is divided according to multiple preset gradient directions based on a preset neighborhood range, resulting in multiple sub-band regions corresponding to each pixel under each preset gradient direction.

[0105] According to an embodiment of the present invention, with each pixel as the center and based on a preset neighborhood range, in the phase interference region image, according to four preset gradient directions of 0°, 45°, 90° and 135°, the region is divided with each pixel as the center, resulting in four sub-band regions corresponding to each pixel under the preset gradient direction of 0°, four sub-band regions corresponding to each pixel under the preset gradient direction of 45°, four sub-band regions corresponding to each pixel under the preset gradient direction of 90° and four sub-band regions corresponding to each pixel under the preset gradient direction of 135°.

[0106] In operation S420, multiple sub-band regions corresponding to each preset gradient direction are weighted to obtain a unidirectional gradient region feature image corresponding to each preset gradient direction.

[0107] According to an embodiment of the present invention, a method for weighting multiple sub-band regions corresponding to each preset gradient direction to obtain a unidirectional gradient region feature image corresponding to each preset gradient direction includes the following operations.

[0108] According to an embodiment of the present invention, for the p-th pixel, where p≥1 and p is a positive integer.

[0109] According to an embodiment of the present invention, the amplitude information in multiple sub-band regions corresponding to each preset gradient direction is enhanced to obtain the detection enhancement gradient value corresponding to each preset gradient direction.

[0110] According to an embodiment of the present invention, a method for enhancing amplitude information in multiple sub-band regions corresponding to each preset gradient direction to obtain a detection enhancement gradient value corresponding to each preset gradient direction includes the following operations.

[0111] According to an embodiment of the present invention, for any preset gradient direction, the amplitude information of each pixel in multiple sub-band regions is averaged to obtain the mean amplitude information of multiple sub-band regions.

[0112] For example, with a preset gradient direction of 0°, four sub-band regions are divided with the (3,5)th pixel in the phase interference region image as the center. Each sub-band region contains 16 pixels. The amplitude information of the 16 pixels in the four sub-band regions is averaged to obtain the average amplitude information of the first sub-band region, the average amplitude information of the second sub-band region, the average amplitude information of the third sub-band region, and the average amplitude information of the fourth sub-band region.

[0113] According to an embodiment of the present invention, based on the sub-band region rule, the difference is calculated according to the mean amplitude information of multiple sub-band regions to obtain the first mean amplitude difference and the second mean amplitude difference.

[0114] According to an embodiment of the present invention, the first mean amplitude difference is calculated from the mean amplitude information of a portion of the sub-band regions in a plurality of sub-band regions, and the second mean amplitude difference is calculated from the mean amplitude information of another portion of the sub-band regions.

[0115] According to an embodiment of the present invention, after dividing the phase interference region pattern based on a preset neighborhood range, the four sub-band regions may include two strong gradient sub-band regions and two weak gradient sub-band regions (the first and second sub-band regions can be strong gradient sub-band regions, and the third and fourth sub-band regions can be weak gradient sub-band regions), with the strong gradient sub-band regions and weak gradient sub-band regions set correspondingly. The sub-band region rule can be characterized as treating one strong gradient sub-band region and one weak gradient sub-band region as a group within the four sub-band regions.

[0116] According to an embodiment of the present invention, generally speaking, the first mean magnitude difference can be obtained by subtracting the mean magnitude information of the first sub-band region of the strong gradient and the mean magnitude information of the third sub-band of the weak gradient, and the second mean magnitude difference can be obtained by subtracting the mean magnitude information of the second sub-band region of the strong gradient and the mean magnitude information of the fourth sub-band region of the weak gradient.

[0117] According to an embodiment of the present invention, the minimum value between the first mean amplitude difference and the second mean amplitude difference is used as the detection enhancement gradient value corresponding to any preset gradient direction.

[0118] According to an embodiment of the present invention, the detection enhancement gradient value can be as shown in formula (4).

[0119] (4);

[0120] in, This can be represented as a coordinate position (x p y pThe pixel detection enhancement gradient value is P1, which can be represented as the mean amplitude information of the first sub-band region, P2, P3, P4, and mean of the second sub-band region. The mean can be represented as the mean amplitude information of the third sub-band region, P4, and mean of the fourth sub-band region.

[0121] According to an embodiment of the present invention, based on the above-described calculation formula (4) for the detection enhancement gradient value, the detection enhancement gradient value corresponding to the 0° preset gradient direction, the detection enhancement gradient value corresponding to the 45° preset gradient direction, the detection enhancement gradient value corresponding to the 90° preset gradient direction, and the detection enhancement gradient value corresponding to the 135° preset gradient direction can be calculated.

[0122] According to an embodiment of the present invention, the standard deviation of the amplitude information in multiple sub-band regions is calculated to obtain a detection weighting operator corresponding to each preset gradient direction.

[0123] According to an embodiment of the present invention, a method for calculating the standard deviation of amplitude information in multiple sub-band regions to obtain a detection weighting operator corresponding to each preset gradient direction includes the following operations.

[0124] According to an embodiment of the present invention, for any preset gradient direction, based on the sub-band region rule, the standard deviation is calculated according to the amplitude information of each pixel in multiple sub-band regions to obtain the first amplitude standard deviation and the second amplitude standard deviation.

[0125] According to an embodiment of the present invention, the first amplitude standard deviation is calculated from the amplitude information of each pixel in a portion of the sub-band region, and the second amplitude standard deviation is calculated from the amplitude information of each pixel in another portion of the sub-band region.

[0126] According to an embodiment of the present invention, the amplitude information of each pixel in each sub-band region can be calculated first. Then, based on the sub-band region rule, the amplitude information of each pixel in the first sub-band region and the amplitude information of each pixel in the third sub-band region of the weak gradient are subtracted to obtain a first amplitude difference value. Then, the standard deviation of the first amplitude difference value is calculated to obtain a first amplitude standard deviation. Similarly, the amplitude information of each pixel in the second sub-band region and the amplitude information of each pixel in the fourth sub-band region of the weak gradient are subtracted to obtain a second amplitude difference value. Then, the standard deviation of the second amplitude difference value is calculated to obtain a second amplitude standard deviation.

[0127] According to an embodiment of the present invention, a detection weighting operator corresponding to any preset gradient direction is obtained based on the first amplitude standard deviation and the second amplitude standard deviation.

[0128] According to an embodiment of the present invention, the detection weighting operator can be as shown in formula (5).

[0129] (5);

[0130] in, This can be represented as a coordinate position (x p y p The pixel detection weighting operator is defined as follows: P1' can be represented as the amplitude information of each pixel in the first sub-band region, P2' can be represented as the amplitude information of each pixel in the second sub-band region, P3' can be represented as the amplitude information of each pixel in the third sub-band region, P4' can be represented as the amplitude information of each pixel in the fourth sub-band region, and std can be represented as the standard deviation processing.

[0131] According to an embodiment of the present invention, based on the above-mentioned calculation formula (5) for the detection weighting operator, the detection weighting operator corresponding to the 0° preset gradient direction, the detection weighting operator corresponding to the 45° preset gradient direction, the detection weighting operator corresponding to the 90° preset gradient direction, and the detection weighting operator corresponding to the 135° preset gradient direction can be calculated.

[0132] According to an embodiment of the present invention, a target detection gradient value corresponding to each preset gradient direction is obtained based on multiple detection enhancement gradient values ​​and multiple detection weighting operators.

[0133] According to an embodiment of the present invention, the target detection gradient value corresponding to the 0° preset gradient direction is calculated for a certain pixel using the detection enhancement gradient value and the detection weighting operator. Similarly, the target detection gradient values ​​corresponding to the 45° preset gradient direction, the 90° preset gradient direction, and the 135° preset gradient direction for a certain pixel can be calculated.

[0134] According to an embodiment of the present invention, the target detection gradient value can be as shown in formula (6).

[0135] (6);

[0136] in, It can be characterized as being related to The target detection gradient value corresponding to the preset gradient direction.

[0137] According to an embodiment of the present invention, a unidirectional gradient region feature image corresponding to each preset gradient direction is constructed based on the target detection gradient values ​​of P pixels corresponding to each preset gradient direction, wherein P ≥ p, and P is a positive integer.

[0138] According to an embodiment of the present invention, there may be a total of P pixels in the phase interference region image. Each pixel in the image is traversed, and the detection enhancement gradient value and detection weighting operator corresponding to each of the P pixels are calculated as described above, thereby obtaining the target detection gradient value of each pixel corresponding to each preset gradient direction. Based on any preset gradient direction, a unidirectional gradient region feature image corresponding to any preset gradient direction is constructed according to the target detection gradient value of each pixel, thereby constructing a unidirectional gradient region feature image corresponding to the 0° preset gradient direction, a 45° preset gradient direction, a 90° preset gradient direction, and a 135° preset gradient direction.

[0139] According to an embodiment of the present invention, for each pixel in the phase interference region image, four sub-band regions are divided with the pixel as the center. Then, for each pixel, based on the mean amplitude information of its corresponding sub-band region and the amplitude information of each pixel, a detection enhancement gradient value corresponding to each preset gradient direction and a detection weighting operator are obtained through minimum value extraction and standard deviation calculation. The detection enhancement gradient value is then weighted and fused using the detection weighting operator, thereby obtaining the target detection gradient value corresponding to each preset gradient direction for that pixel. The above-mentioned process of weighting and fusing the detection enhancement gradient value with each preset gradient direction is performed on each pixel in the phase interference region image. The algorithm processes and calculates the target detection gradient value corresponding to the gradient direction. Then, based on each preset gradient direction, it constructs a unidirectional gradient region feature image under the same preset gradient direction according to the target detection gradient value of each pixel under the same preset gradient direction. This realizes the construction of the unidirectional gradient region feature image. By preset gradient directions of 0°, 45°, 90° and 135°, it fully explores the unidirectional gradient features of small foreign objects at various angles of the spatial structure. While suppressing the interference of high brightness pixel noise, it enhances the signal-to-noise ratio of weak targets, so as to facilitate the accurate detection of small foreign object fragments.

[0140] Figure 5 A schematic diagram of multiple sub-band regions with a preset gradient direction of 0° is shown according to an embodiment of the present invention.

[0141] like Figure 5 As shown, Figure 5 As shown, a triangle can be represented by any pixel in the phase interference region image. Based on the preset neighborhood range and the preset gradient direction of 0°, the region is divided to obtain the first sub-band region P1, the second sub-band region P2, the third sub-band region P3, and the fourth sub-band region P4 under the preset gradient direction of 0°.

[0142] Figure 6 A schematic diagram of multiple sub-band regions with a preset gradient direction of 45° is shown according to an embodiment of the present invention.

[0143] like Figure 6 As shown, Figure 6 As shown, a triangle can be represented by any pixel in the phase interference region image. Centered on this pixel, the region is divided based on a preset neighborhood range and a preset gradient direction of 45° to obtain the first sub-band region P1, the second sub-band region P2, the third sub-band region P3, and the fourth sub-band region P4 along the preset gradient direction of 45°.

[0144] Figure 7 A schematic diagram of multiple sub-band regions with a preset gradient direction of 90° is shown according to an embodiment of the present invention.

[0145] like Figure 7 As shown, Figure 7 As shown, a triangle can be represented by any pixel in the phase interference region image. Centered on this pixel, the region is divided based on a preset neighborhood range and a preset gradient direction of 90° to obtain the first sub-band region P1, the second sub-band region P2, the third sub-band region P3, and the fourth sub-band region P4 along the preset gradient direction of 90°.

[0146] Figure 8 A schematic diagram of multiple sub-band regions with a preset gradient direction of 135° is shown according to an embodiment of the present invention.

[0147] like Figure 8 As shown, Figure 8 As shown, a triangle can be represented by any pixel in the phase interference region image. Centered on this pixel, the region is divided based on a preset neighborhood range and a preset gradient direction of 135° to obtain the first sub-band region P1, the second sub-band region P2, the third sub-band region P3, and the fourth sub-band region P4 along the preset gradient direction of 135°.

[0148] In operation S430, multiple unidirectional gradient region feature images are multiplied together to obtain multidirectional gradient region feature images.

[0149] According to an embodiment of the present invention, the multi-directional gradient region feature image can be as shown in formula (7).

[0150] (7);

[0151] Among them, C f It can be characterized as a multi-directional gradient region feature image. It can be characterized as being related to Feature image of the unidirectional gradient region corresponding to the preset gradient direction.

[0152] According to an embodiment of the present invention, the phase interference region image is divided according to multiple preset gradient directions based on a preset neighborhood range to obtain multiple sub-band regions corresponding to each pixel under each preset gradient direction. Then, the multiple sub-band regions corresponding to each preset gradient direction are weighted to obtain a unidirectional gradient region feature image corresponding to each preset gradient direction. Then, the multiple unidirectional gradient region feature images are multiplied to obtain a multidirectional gradient region feature image. This realizes the multi-angle feature extraction of each pixel in the phase interference region image after multiple iterations of registration adjustment. By calculating the enhancement value and the weight used for weighting, the enhancement features are weighted by standard deviation to construct a unidirectional gradient region feature image corresponding to each spatial angle. Then, the multiple unidirectional gradient region feature images are fused into multidirectional features. Even in complex environments, accurate target detection results can be obtained through multi-dimensional feature analysis, improving detection accuracy, reducing the probability of missed detections and false alarms, and improving detection efficiency.

[0153] According to an embodiment of the present invention, a method for dividing a multi-directional gradient region feature image based on a first gradient threshold to obtain a target detection result corresponding to the target region includes the following operations.

[0154] According to an embodiment of the present invention, based on a first gradient threshold, the feature image of the multi-directional gradient region is segmented to obtain target detection gradient points greater than the first gradient threshold.

[0155] According to an embodiment of the present invention, based on a preset first gradient threshold, the feature image of the multi-directional gradient region is judged. If a gradient feature point greater than or equal to the first gradient threshold appears in the feature image of the multi-directional gradient region, the presence of a small foreign object fragment is confirmed and a target detection gradient point is obtained. If no gradient feature point greater than or equal to the first gradient threshold appears in the feature image of the multi-directional gradient region, the absence of a small foreign object fragment is confirmed.

[0156] According to an embodiment of the present invention, a target detection result corresponding to the target region is obtained based on the target detection gradient point and the detection coordinate information of the target detection gradient point.

[0157] According to an embodiment of the present invention, depending on whether a target detection gradient point exists, and if a target detection gradient point exists, a target detection result can be generated based on the azimuth information and range information corresponding to the target detection gradient point.

[0158] According to an embodiment of the present invention, by dividing the feature image of the multi-directional gradient region based on a first gradient threshold, target detection gradient points greater than the first gradient threshold are obtained. Then, based on the target detection gradient points and their detection coordinate information, the target detection result corresponding to the target region is obtained. This achieves accurate detection of tiny foreign object fragments by setting a first gradient threshold and based on the features extracted from the multi-dimensional spatial angle in the feature image of the multi-directional gradient region, thereby improving detection accuracy, reducing the probability of missed detections and false alarms, and improving detection efficiency.

[0159] Figure 9 A schematic diagram of a multi-directional gradient region feature image according to an embodiment of the present invention is shown.

[0160] like Figure 9 As shown, Figure 9 The image shows a multi-directional gradient region feature image. The x-axis of the multi-directional gradient region feature image can be represented by the azimuth direction, the y-axis by the range direction, and the z-axis by the gradient intensity. With the first gradient threshold of -200, it can be seen from the image that there are four target detection gradient points. The first target detection gradient point is located at -0.75m from the radar azimuth direction and 8m from the range direction. The second target detection gradient point is located at -0.7m from the radar azimuth direction and 10m from the range direction. The third target detection gradient point is located at 0.2m from the radar azimuth direction and 10m from the range direction. The first target detection gradient point is located at 0.25m from the radar azimuth direction and 11m from the range direction.

[0161] Figure 10 A structural block diagram of a radar detection device based on multi-directional gradient enhancement detection according to an embodiment of the present invention is shown.

[0162] like Figure 10 As shown, the radar detection device based on multi-directional gradient enhancement detection in this embodiment includes: an imaging module 1010, a registration module 1020, a gradient processing module 1030, and a segmentation module 1040.

[0163] The imaging module 1010 is used to perform radar imaging processing on the electromagnetic echo signal reflected by the target area to be detected in response to receiving the electromagnetic echo signal, thereby obtaining a detection area image of the target area. The imaging module 1010 can be used to perform the operation S210 described above, which will not be repeated here.

[0164] The registration module 1020 is used to register the detection region image and the reference region image corresponding to the target region to obtain the phase interference region image of the target region. The registration module 1020 can be used to perform the operation S220 described above, which will not be repeated here.

[0165] The gradient processing module 1030 is used to perform gradient processing on multiple pixels in the phase interference region image based on a preset neighborhood range and multiple preset gradient directions to obtain a multi-directional gradient region feature image. This multi-directional gradient region feature image represents the image obtained after fusing it with the image features corresponding to each preset gradient direction. The gradient processing module 1030 can be used to perform the operation S230 described above, which will not be repeated here.

[0166] The segmentation module 1040 is used to segment the feature image of the multi-directional gradient region based on the first gradient threshold to obtain the target detection result corresponding to the target region. The segmentation module 1040 can be used to perform the operation S240 described above, which will not be repeated here.

[0167] According to an embodiment of the present invention, the registration module 1020 includes: a first determining submodule, a first constructing submodule, and a first adjusting submodule.

[0168] The first determination submodule is used to determine the detection feature window and the reference feature window from the detection region image and the reference region image, respectively, based on feature segmentation rules.

[0169] The first construction submodule is used to construct a feature cross-correlation function by processing the detection coordinate information and detection amplitude information of multiple detection feature pixels in the detection feature window and the reference coordinate information and reference amplitude information of multiple reference feature pixels in the reference feature window.

[0170] The first adjustment submodule is used to iteratively adjust the detection region image based on the feature cross-correlation function to obtain the phase interference region image.

[0171] According to an embodiment of the present invention, the first adjustment submodule includes: a first iterative adjustment unit, a first execution unit, and a first adjustment unit.

[0172] The first iterative adjustment unit is used to adjust the adjusted detection image of the (i-1)th iteration round for the i-th iteration round, where i ≥ 2 and i is a positive integer; using the (i-1)th first offset information determined from the (i-1)th feature cross-correlation function, to obtain the adjusted detection image of the i-th iteration round, where the first offset information represents the position information corresponding to the feature cross-correlation peak in the feature cross-correlation function; based on the feature partitioning rules, the i-th detection feature window and the i-th reference feature window are determined from the adjusted detection image and the reference region image of the i-th iteration round, respectively; by processing the detection coordinate information and detection amplitude information of multiple detection feature pixels in the i-th detection feature window and the reference coordinate information and reference amplitude information of multiple reference feature pixels in the i-th reference feature window, the i-th feature cross-correlation function is constructed; and the i-th first offset information corresponding to the i-th feature cross-correlation peak is determined from the i-th feature cross-correlation function.

[0173] The first execution unit is configured to repeatedly execute the above operation until the i-th first offset information is less than the predetermined offset threshold when the i-th first offset information is greater than or equal to the predetermined offset threshold.

[0174] The first adjustment unit is used to adjust the adjustment detection image of the i-th iteration using the i-th first offset information when the i-th first offset information is less than a predetermined offset threshold, so as to obtain a phase interference region image.

[0175] According to an embodiment of the present invention, the gradient processing module 1030 includes: a first partitioning submodule, a first weighting submodule, and a first accumulating submodule.

[0176] The first partitioning submodule is used to partition the phase interference region image based on a preset neighborhood range and multiple preset gradient directions to obtain multiple sub-band regions corresponding to each pixel under each preset gradient direction.

[0177] The first weighting submodule is used to perform weighted processing on multiple sub-band regions corresponding to each preset gradient direction to obtain a unidirectional gradient region feature image corresponding to each preset gradient direction.

[0178] The first cumulative multiplication submodule is used to perform cumulative multiplication on multiple unidirectional gradient region feature images to obtain multidirectional gradient region feature images.

[0179] According to an embodiment of the present invention, the first weighting submodule further includes: a first weighting unit and a first construction unit.

[0180] The first weighting unit is used to perform amplitude enhancement processing on the amplitude information in multiple sub-band regions corresponding to each preset gradient direction for the p-th pixel, where p≥1 and p is a positive integer, to obtain the detection enhancement gradient value corresponding to each preset gradient direction; calculate the standard deviation of the amplitude information in multiple sub-band regions to obtain the detection weighting operator corresponding to each preset gradient direction; and obtain the target detection gradient value corresponding to each preset gradient direction based on the multiple detection enhancement gradient values ​​and the multiple detection weighting operators.

[0181] The first construction unit is used to construct a unidirectional gradient region feature image corresponding to each preset gradient direction based on the target detection gradient values ​​of P pixels corresponding to each preset gradient direction, where P ≥ p and P is a positive integer.

[0182] According to an embodiment of the present invention, the first weighting unit further includes: a first averaging subunit, a first difference subunit, and a first determination subunit.

[0183] The first averaging sub-unit is used to average the amplitude information of each pixel in multiple sub-band regions for any preset gradient direction, so as to obtain the mean amplitude information of multiple sub-band regions.

[0184] The first difference sub-unit is used to calculate the difference based on the mean amplitude information of multiple sub-band regions according to the sub-band region rules, to obtain the first mean amplitude difference and the second mean amplitude difference. The first mean amplitude difference is calculated from the mean amplitude information of some sub-band regions among the multiple sub-band regions, and the second mean amplitude difference is calculated from the mean amplitude information of another part of the sub-band regions.

[0185] The first determining subunit is used to take the minimum value between the first mean amplitude difference and the second mean amplitude difference as the detection enhancement gradient value corresponding to any preset gradient direction.

[0186] According to an embodiment of the present invention, the first weighting unit further includes: a first standard deviation subunit and a first obtaining subunit.

[0187] The first standard deviation sub-unit is used to calculate the standard deviation for any preset gradient direction based on the sub-band region rule and the amplitude information of each pixel in multiple sub-band regions, to obtain the first amplitude standard deviation and the second amplitude standard deviation. The first amplitude standard deviation is calculated from the amplitude information of each pixel in a portion of the sub-band regions, and the second amplitude standard deviation is calculated from the amplitude information of each pixel in a portion of the sub-band regions.

[0188] The first sub-unit is used to obtain a detection weighting operator corresponding to any preset gradient direction based on the first amplitude standard deviation and the second amplitude standard deviation.

[0189] According to an embodiment of the present invention, the partitioning module 1040 further includes a second partitioning submodule and a first obtaining submodule.

[0190] The second segmentation submodule is used to segment the feature image of the multi-directional gradient region based on the first gradient threshold to obtain target detection gradient points that are greater than the first gradient threshold.

[0191] The first submodule is used to obtain the target detection result corresponding to the target region based on the target detection gradient point and the detection coordinate information of the target detection gradient point.

[0192] According to embodiments of the present invention, any plurality of modules among the imaging module 1010, registration module 1020, gradient processing module 1030, and partitioning module 1040 may be combined into one module, or any one of these modules may be split into multiple modules. Alternatively, at least a portion of the functionality of one or more of these modules may be combined with at least a portion of the functionality of other modules and implemented in one module. According to embodiments of the present invention, at least one of the imaging module 1010, registration module 1020, gradient processing module 1030, and partitioning module 1040 may be at least partially implemented as hardware circuitry, such as a field-programmable gate array (FPGA), a programmable logic array (PLA), a system-on-a-chip, a system-on-a-substrate, a system-on-package, an application-specific integrated circuit (ASIC), or any other reasonable means of integrating or packaging the circuitry, or implemented in software, hardware, or firmware, or in any suitable combination of any of these three implementation methods. Alternatively, at least one of the imaging module 1010, registration module 1020, gradient processing module 1030, and segmentation module 1040 may be implemented at least partially as a computer program module, which can perform corresponding functions when the computer program module is run.

[0193] Figure 11 A block diagram of an electronic device based on a radar detection method using multi-directional gradient enhancement detection according to an embodiment of the present invention is shown.

[0194] like Figure 11As shown, an electronic device according to an embodiment of the present invention includes a processor 1101, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1102 or a program loaded from a storage portion 1108 into a random access memory (RAM) 1103. The processor 1101 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or an associated chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 1101 may also include onboard memory for caching purposes. The processor 1101 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of the present invention.

[0195] RAM 1103 stores various programs and data required for the operation of the electronic device. Processor 1101, ROM 1102, and RAM 1103 are interconnected via bus 1104. Processor 1101 executes various operations of the method flow according to embodiments of the present invention by executing programs in ROM 1102 and / or RAM 1103. It should be noted that the programs may also be stored in one or more memories other than ROM 1102 and RAM 1103. Processor 1101 may also execute various operations of the method flow according to embodiments of the present invention by executing programs stored in said one or more memories.

[0196] According to embodiments of the present invention, the electronic device may further include an input / output (I / O) interface 1105, which is also connected to a bus 1104. The electronic device may also include one or more of the following components connected to the I / O interface 1105: an input section 1106 including a keyboard, mouse, etc.; an output section 1107 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 1108 including a hard disk, etc.; and a communication section 1109 including a network interface card such as a LAN card, modem, etc. The communication section 1109 performs communication processing via a network such as the Internet. A drive 1110 is also connected to the I / O interface 1105 as needed. A removable medium 1111, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on the drive 1110 as needed so that computer programs read from it can be installed into the storage section 1108 as needed.

[0197] The present invention also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments; or it may exist independently and not assembled into the device / apparatus / system. The computer-readable storage medium carries one or more programs, which, when executed, implement the method according to the embodiments of the present invention.

[0198] According to embodiments of the present invention, a computer-readable storage medium may be a non-volatile computer-readable storage medium, such as including, but not limited to: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. For example, according to embodiments of the present invention, a computer-readable storage medium may include ROM 1102 and / or RAM 1103 and / or one or more memories other than ROM 1102 and RAM 1103 described above.

[0199] Embodiments of the present invention also include a computer program product comprising a computer program containing program code for performing the methods shown in the flowchart. When the computer program product is run on a computer system, the program code is used to enable the computer system to implement the radar detection method based on multi-directional gradient enhancement detection provided in the embodiments of the present invention.

[0200] When the computer program is executed by the processor 1101, it performs the functions defined in the system / apparatus of this embodiment of the invention. According to embodiments of the invention, the systems, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0201] In one embodiment, the computer program may rely on a tangible storage medium such as an optical storage device or a magnetic storage device. In another embodiment, the computer program may also be transmitted and distributed in the form of signals over a network medium, and may be downloaded and installed via the communication section 1109, and / or installed from the removable medium 1111. The program code contained in the computer program can be transmitted using any suitable network medium, including but not limited to: wireless, wired, etc., or any suitable combination thereof.

[0202] In such an embodiment, the computer program can be downloaded and installed from a network via communication section 1109, and / or installed from removable medium 1111. When the computer program is executed by processor 1101, it performs the functions defined in the system of this embodiment of the invention. According to embodiments of the invention, the systems, devices, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0203] Those skilled in the art will understand that the features described in the various embodiments of the present invention can be combined or combined in various ways, even if such combinations or combinations are not explicitly described in the present invention. In particular, the features described in the various embodiments of the present invention can be combined and / or combined in various ways without departing from the spirit and teachings of the present invention. All such combinations and / or combinations fall within the scope of the present invention.

Claims

1. A radar detection method based on multi-directional gradient enhancement detection, characterized in that, include: In response to receiving an electromagnetic echo signal reflected by a target area to be detected, radar imaging processing is performed on the electromagnetic echo signal to obtain a detection area image of the target area; The detection region image and the reference region image corresponding to the target region are registered to obtain the phase interference region image of the target region. Based on a preset neighborhood range, gradient processing is performed on multiple pixels in the phase interference region image according to multiple preset gradient directions to obtain a multi-directional gradient region feature image. The multi-directional gradient region feature image represents the image obtained by fusing image features corresponding to each preset gradient direction. Obtaining the multi-directional gradient region feature image includes: dividing the phase interference region image according to the multiple preset gradient directions based on the preset neighborhood range to obtain multiple sub-band regions corresponding to each pixel under each preset gradient direction; weighting the multiple sub-band regions corresponding to each preset gradient direction to obtain a unidirectional gradient region feature image corresponding to each preset gradient direction; and multiplying the multiple unidirectional gradient region feature images to obtain the multi-directional gradient region feature image. Based on the first gradient threshold, the feature image of the multi-directional gradient region is segmented to obtain the target detection result corresponding to the target region.

2. The method according to claim 1, characterized in that, The step of registering the detection region image and the reference region image corresponding to the target region to obtain the phase interference region image of the target region includes: Based on feature segmentation rules, a detection feature window and a reference feature window are determined from the detection region image and the reference region image, respectively. By processing the detection coordinate information and detection amplitude information of multiple detection feature pixels in the detection feature window and the reference coordinate information and reference amplitude information of multiple reference feature pixels in the reference feature window, a feature cross-correlation function is constructed. Based on the feature cross-correlation function, the detection region image is iteratively adjusted to obtain the phase interference region image.

3. The method according to claim 2, characterized in that, The step of iteratively adjusting the detection region image based on the feature cross-correlation function to obtain the phase interference region image includes: For the i-th iteration, where i ≥ 2, i is a positive integer; Using the first offset information determined from the (i-1)th feature cross-correlation function, the adjusted detection image of the (i-1)th iteration round is adjusted to obtain the adjusted detection image of the ith iteration round. Here, the first offset information is characterized as the position information corresponding to the feature cross-correlation peak in the feature cross-correlation function. Based on the feature segmentation rules, the i-th detection feature window and the i-th reference feature window are determined from the adjusted detection image of the i-th iteration and the reference region image, respectively; By processing the detection coordinate information and detection amplitude information of multiple detection feature pixels in the i-th detection feature window and the reference coordinate information and reference amplitude information of multiple reference feature pixels in the i-th reference feature window, the i-th feature cross-correlation function is constructed. Determine the i-th first offset information corresponding to the peak value of the i-th feature cross-correlation function; If the i-th first offset information is greater than or equal to a predetermined offset threshold, the above operation is repeated until the i-th first offset information is less than the predetermined offset threshold; If the i-th first offset information is less than the predetermined offset threshold, the i-th first offset information is used to adjust the adjustment detection image of the i-th iteration to obtain the phase interference region image.

4. The method according to claim 1, characterized in that, The step of weighting multiple sub-band regions corresponding to each preset gradient direction to obtain a unidirectional gradient region feature image corresponding to each preset gradient direction includes: For the p-th pixel, where p≥1 and p is a positive integer; The amplitude information in the plurality of sub-band regions corresponding to each preset gradient direction is enhanced to obtain the detection enhancement gradient value corresponding to each preset gradient direction; The standard deviation of the amplitude information in the multiple sub-band regions is calculated to obtain the detection weighting operator corresponding to each preset gradient direction; Based on multiple detection enhancement gradient values ​​and multiple detection weighting operators, the target detection gradient value corresponding to each preset gradient direction is obtained; Based on the target detection gradient values ​​of P pixels corresponding to each preset gradient direction, a unidirectional gradient region feature image corresponding to each preset gradient direction is constructed, where P ≥ p and P is a positive integer.

5. The method according to claim 4, characterized in that, The step of enhancing the amplitude information in the plurality of sub-band regions corresponding to each preset gradient direction to obtain the detection enhancement gradient value corresponding to each preset gradient direction includes: For any preset gradient direction, the amplitude information of each pixel in the plurality of sub-band regions is averaged to obtain the mean amplitude information of the plurality of sub-band regions. Based on the sub-band region rules, the difference is calculated according to the mean amplitude information of the multiple sub-band regions to obtain a first mean amplitude difference and a second mean amplitude difference. The first mean amplitude difference is calculated from the mean amplitude information of a portion of the multiple sub-band regions, and the second mean amplitude difference is calculated from the mean amplitude information of another portion of the sub-band regions. The minimum value between the first mean amplitude difference and the second mean amplitude difference is taken as the detection enhancement gradient value corresponding to any preset gradient direction.

6. The method according to claim 4, characterized in that, The step of calculating the standard deviation of the amplitude information in the plurality of sub-band regions to obtain a detection weighting operator corresponding to each preset gradient direction includes: For any preset gradient direction, based on the sub-band region rule, the standard deviation is calculated according to the amplitude information of each pixel in the multiple sub-band regions to obtain a first amplitude standard deviation and a second amplitude standard deviation. The first amplitude standard deviation is calculated from the amplitude information of each pixel in a portion of the sub-band regions, and the second amplitude standard deviation is calculated from the amplitude information of each pixel in another portion of the sub-band regions. Based on the first amplitude standard deviation and the second amplitude standard deviation, the detection weighting operator corresponding to any preset gradient direction is obtained.

7. The method according to claim 1, characterized in that, The step of segmenting the multi-directional gradient region feature image based on a first gradient threshold to obtain a target detection result corresponding to the target region includes: Based on the first gradient threshold, the feature image of the multi-directional gradient region is divided to obtain target detection gradient points that are greater than the first gradient threshold. Based on the target detection gradient points and their detection coordinates, a target detection result corresponding to the target region is obtained.

8. A radar detection device based on multi-directional gradient enhancement detection, characterized in that, include: An imaging module is used to respond to receiving an electromagnetic echo signal reflected by a target area to be detected, perform radar imaging processing on the electromagnetic echo signal, and obtain a detection area image of the target area. The registration module is used to perform registration processing on the detection region image and the reference region image corresponding to the target region to obtain the phase interference region image of the target region. A gradient processing module is used to perform gradient processing on multiple pixels in the phase interference region image based on a preset neighborhood range and multiple preset gradient directions to obtain a multi-directional gradient region feature image. The multi-directional gradient region feature image represents an image obtained by fusing image features corresponding to each preset gradient direction. Obtaining the multi-directional gradient region feature image includes: dividing the phase interference region image based on the preset neighborhood range and multiple preset gradient directions to obtain multiple sub-band regions corresponding to each pixel under each preset gradient direction; performing weighted processing on the multiple sub-band regions corresponding to each preset gradient direction to obtain a unidirectional gradient region feature image corresponding to each preset gradient direction; and performing cumulative multiplication processing on the multiple unidirectional gradient region feature images to obtain the multi-directional gradient region feature image. The segmentation module is used to segment the feature image of the multi-directional gradient region based on a first gradient threshold to obtain the target detection result corresponding to the target region.

9. An electronic device, characterized in that, include: One or more processors; Storage device for storing one or more programs. Wherein, when the one or more programs are executed by the one or more processors, the one or more processors perform the method according to any one of claims 1 to 7.

Images