Scene matching method and scene matching system based on visual navigation of unmanned aerial vehicle

Abstract

The invention relates to a scene matching method and a scene matching system based on unmanned aerial vehicle visual navigation, and the method comprises the steps: obtaining an image in a coverage area, and carrying out the segmentation processing of the obtained image, and obtaining a segmented image; selecting one segmented image as a positioning reference image; analyzing and processing the positioning reference image to obtain a feature parameter set; retrieving in the positioning reference image by using the feature parameters to obtain a plurality of regions of interest; and determining an overlapped part of the plurality of regions of interest, recording the overlapped part as a suspected position region, and screening the suspected position region by using the feature parameter set to obtain a current position region. According to the scene matching method and the scene matching system based on the unmanned aerial vehicle visual navigation disclosed by the invention, the suspected multi-position screening and the accurate position determination are carried out by carrying out segmentation processing on the image and combining a detail feature comparison mode, and the mode can improve the mirror image matching speed and meet the positioning requirement of rapid movement of the unmanned aerial vehicle.

Description

Technical Field

[0001] The present application relates to the field of autonomous navigation technology, and in particular to a scene matching method and a scene matching system based on unmanned aerial vehicle visual navigation. Background Art

[0002] Drone visual navigation is mainly based on computer vision technology, which realizes autonomous navigation and positioning of drones by processing and analyzing image information captured by cameras. The camera on the drone acts as its "eyes", while the image processing algorithm acts as the "brain" to screen, analyze and judge visual information.

[0003] In actual application scenarios, an important branch technology in UAV visual navigation is scene matching navigation, which can achieve precise positioning of UAVs in a wider environment. The specific implementation method is to match the real-time captured images with the basic three-dimensional map during the flight to calculate its own position.

[0004] One of the current technical difficulties is that in dynamically updated scenes, drones rely on mirror matching for positioning. However, when the scene matching speed cannot meet the drone's movement speed requirements, it is easy to cause the drone's positioning to be inaccurate and its movement speed to be limited. Summary of the invention

[0005] The present application provides a scene matching method and a scene matching system based on UAV visual navigation, which performs suspected multi-position screening and precise position determination by segmenting the image and combining it with detail feature comparison. This method can improve the mirror matching speed and meet the positioning needs of fast-moving UAVs.

[0006] The above-mentioned purpose of the present application is achieved through the following technical solutions: In a first aspect, the present application provides a scene matching method based on UAV visual navigation, comprising: Acquire an image within the coverage area and perform segmentation processing on the acquired image to obtain a segmented image; Select a segmented image as a positioning reference image; Analyze the positioning reference image and obtain a feature parameter set, where the feature parameter set includes multiple feature parameters; Using the characteristic parameters to search in the positioning reference image, a region of interest is obtained, and the number of the region of interest is multiple; Determine the overlapping parts of multiple interest regions and record them as suspected location regions; The suspected location area is screened using the feature parameter set to obtain the current location area.

[0007] In a possible implementation manner of the first aspect, selecting a segmented image as a positioning reference image includes: Perform grayscale processing on the segmented image to obtain a grayscale segmented image; Extracting local detail features in the grayscale segmentation image and evaluating the local detail features to obtain evaluation results; Use the evaluation results to sort the grayscale segmented images; The segmented image corresponding to the first grayscale segmented image in the sequential sequence is selected as the positioning reference image.

[0008] In a possible implementation manner of the first aspect, extracting local detail features in the grayscale segmented image and evaluating the local detail features includes: Identify feature objects included in the grayscale segmented image, where the feature objects include moving objects and non-moving objects; Construct a reference grid using moving and non-moving objects; Adjust the reference grid so that the number of edges of the grids in the reference grid are consistent; The number of grid nodes and the influence points between adjacent grid nodes are weighted to obtain the evaluation results.

[0009] In a possible implementation manner of the first aspect, parsing the positioning reference image to obtain a feature parameter set includes: Extracting characteristic objects in the positioning reference image, the characteristic objects include moving objects and non-moving objects; Assign type labels to trait objects; Extract local line segment features of feature objects, including shape and color change direction; Calculate the distance value of adjacent local line segment features; The type label, the local line segment feature of the feature object and the distance value of the adjacent local line segment feature are taken as a feature parameter set.

[0010] In a possible implementation manner of the first aspect, using the feature parameters to search in the positioning reference image and obtain the region of interest includes: Use the type label of the feature object to search in the positioning reference image to obtain the suspected area of ​​interest; The local line segment features of the feature object are used to screen the suspected interest regions to obtain the interest regions.

[0011] In a possible implementation manner of the first aspect, screening the suspected interest region by using the local line segment feature of the feature object includes: Determine the grayscale value range of the local line segment feature of the feature object and generate a grayscale extraction interval according to the grayscale value range; The suspected region of interest is extracted using the grayscale extraction interval to obtain an extracted image; Use the local line segment features of the feature object to traverse in the extracted image; When there are local line segment features of the feature object in the extracted image, the suspected interest region corresponding to the extracted image is retained, otherwise the suspected interest region corresponding to the extracted image is deleted.

[0012] In a possible implementation manner of the first aspect, using the local line segment features of the feature object to traverse the extracted image further includes: Sequentially divide the grayscale extraction interval into a plurality of sub-grayscale extraction intervals; The local line segment features and suspected interest regions of the feature objects are extracted using the same sub-grayscale extraction interval; Determine whether there is a local line segment feature of a feature object extracted in the same sub-grayscale extraction interval in the extracted image extracted in the same sub-grayscale extraction interval, and obtain an existence result; Determine whether there are local line segment features of the feature object in the extracted image based on the existence result.

[0013] In a second aspect, the present application provides a scene matching device based on drone visual navigation, comprising: An image processing unit is used to acquire an image in the coverage area and segment the acquired image to obtain a segmented image; An image selection unit, used for selecting a segmented image as a positioning reference image; An image parsing unit, used to parse the positioning reference image and obtain a feature parameter set, wherein the feature parameter set includes a plurality of feature parameters; A retrieval processing unit, used to use the characteristic parameters to search in the positioning reference image to obtain a region of interest, where the number of the regions of interest is multiple; A first position determination unit is used to determine an overlapping portion of multiple interest regions and record it as a suspected position region; The second location determination unit is used to screen the suspected location area using the feature parameter set to obtain the current location area.

[0014] In a third aspect, the present application provides a scene matching system based on drone visual navigation, the system comprising: one or more memories for storing instructions; and One or more processors, used to call and run the instructions from the memory to execute the method as described in the first aspect and any possible implementation of the first aspect.

[0015] In a fourth aspect, the present application provides a computer-readable storage medium, the computer-readable storage medium comprising: Program, when the program is executed by a processor, the method described in the first aspect and any possible implementation of the first aspect is executed.

[0016] In a fifth aspect, the present application provides a computer program product, comprising program instructions. When the program instructions are executed by a computing device, the method described in the first aspect and any possible implementation of the first aspect is executed.

[0017] In a sixth aspect, the present application provides a chip system, which includes a processor for implementing the functions involved in the above aspects, for example, generating, receiving, sending, or processing the data and / or information involved in the above methods.

[0018] The chip system may be composed of chips, or may include chips and other discrete devices.

[0019] In a possible design, the chip system also includes a memory, which is used to store necessary program instructions and data. The processor and the memory can be decoupled and respectively set on different devices, connected by wired or wireless means, or the processor and the memory can also be coupled on the same device. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 This is a schematic diagram of the steps of a scene matching method based on UAV visual navigation provided in this application.

[0021] Figure 2 This is a schematic diagram of the principle of achieving positioning by scene matching provided by this application.

[0022] Figure 3 This is a schematic diagram of segmenting an acquired image provided by the present application.

[0023] Figure 4 It is a schematic diagram of constructing a reference grid provided in this application. DETAILED DESCRIPTION

[0024] The technical solution in this application is further described in detail below in conjunction with the accompanying drawings.

[0025] This application discloses a scene matching method based on UAV visual navigation, please refer to Figure 1 The scene matching method based on UAV visual navigation disclosed in this application includes the following steps: S101, acquiring an image within a coverage area and performing segmentation processing on the acquired image to obtain a segmented image; S102, selecting a segmented image as a positioning reference image; S103, analyzing and processing the positioning reference image to obtain a feature parameter set, where the feature parameter set includes a plurality of feature parameters; S104, searching in the positioning reference image using the feature parameters to obtain a region of interest, where the number of the regions of interest is multiple; S105, determining overlapping parts of multiple interest regions and recording them as suspected location regions; S106, using the characteristic parameter set to filter the suspected location area to obtain the current location area.

[0026] The scene matching method based on UAV visual navigation disclosed in the present application is applied to the use of UAV in scenarios where satellite positioning cannot be used. The scenarios where satellite positioning cannot be used include situations where the satellite positioning signal is weak and the satellite positioning signal is interfered with.

[0027] When the drone cannot use satellite positioning to locate, it needs to use scene matching to determine the current position, such as Figure 2 As shown in the figure, the large rectangle represents the image in the coverage area, and the small rectangle in the lower left corner represents the image taken by the drone. When the small rectangle successfully matches the dotted rectangle inside the large rectangle, the current coordinates can be obtained according to the coordinates corresponding to the dotted rectangle.

[0028] In the specific matching method, it is first necessary to obtain an image within the coverage area and segment the obtained image to obtain a segmented image, that is, the content in step S101, such as Figure 3 shown.

[0029] There are two reasons for segmenting the image. One is that the coverage area is large, resulting in the image containing too much content. The other is that the quality of the content contained in the image is uneven and needs to be screened.

[0030] Next, in step S102, a segmented image is selected as a positioning reference image. After the positioning reference image is obtained, the positioning reference image is analyzed and processed to obtain a feature parameter set, which is the content in step S103. The feature parameter set consists of multiple feature parameters.

[0031] Then, in step S104, the feature parameters are used to search in the positioning reference image to obtain the region of interest, and the number of the region of interest is multiple. Here, because the number of feature parameters is multiple, when the feature parameters are used to search in the positioning reference image, each feature parameter can generate a region of interest, and the difference lies in the size or number of the region of interest.

[0032] Of course, there may be a situation where a certain feature parameter does not obtain the region of interest, but the probability of this situation occurring is extremely small, and the application of multiple feature parameters can minimize or even eliminate the negative impact caused by this situation.

[0033] In step S105 and step S106, the overlapping parts of multiple interest regions are first determined and recorded as suspected location regions, and then the suspected location regions are screened using a feature parameter set to obtain a current location region.

[0034] Here, when determining the overlapping parts of multiple regions of interest, the overlapping parts of the regions of interest can also be sorted according to the number of overlaps in the overlapping parts, and then the overlapping parts of the first few regions of interest in the sequential sequence are retained as suspected location areas according to the sorting results, and finally the suspected location areas are screened using a set of feature parameters.

[0035] This method can significantly reduce the number of suspected location areas.

[0036] In some examples, a segmented image is selected as a positioning reference image in the following manner: S201, performing grayscale processing on the segmented image to obtain a grayscale segmented image; S202, extracting local detail features in the grayscale segmented image and evaluating the local detail features to obtain an evaluation result; S203, sorting the grayscale segmented images using the evaluation results; S204, selecting a segmented image corresponding to the first grayscale segmented image in the sequential sequence as a positioning reference image.

[0037] In the contents of step S201 to step S204, the segmented image needs to be gray-scaled first and then evaluated. The purpose of gray-scale processing is to reduce the amount of data processing. Then, local detail features in the gray-scale segmented image are extracted and evaluated.

[0038] The evaluation method is to obtain a specific value based on the extracted local detail features, and then sort multiple grayscale segmentation images according to the value, and then select the segmentation image corresponding to the first grayscale segmentation image in the sequential sequence as the positioning reference image.

[0039] The specific method of extracting local detail features in the grayscale segmentation image and evaluating the local detail features is as follows: S301, identifying feature objects included in the grayscale segmented image, where the feature objects include moving objects and non-moving objects; S302, constructing a reference grid using moving objects and non-moving objects; S303, adjusting the reference grid so that the number of edges of the grids in the reference grid are consistent; S304, performing weighted calculation on the number of grid nodes and the influence points between adjacent grid nodes to obtain an evaluation result.

[0040] The mobile objects in step S301 refer to vehicles, personnel, and mechanical equipment, etc., and the non-mobile objects refer to fixed attachments on the ground. Then, the mobile objects and non-mobile objects are used to construct a reference grid, and the reference grid is adjusted so that the number of edges of the grids in the reference grid is consistent.

[0041] Then, the moving objects and non-moving objects are used to construct a reference grid and the number of edges of the grids in the reference grid is adjusted so that the number of edges of each grid is the same, for example, all are triangles or quadrilaterals. Finally, the number of grid nodes and the influence points between adjacent grid nodes are weighted to obtain the evaluation results.

[0042] From the above content, it can be concluded that the evaluation results include two indicators, namely the number of grid nodes and the influence points between adjacent grid nodes. The number of grid nodes refers to the number of feature objects involved in constructing the reference grid, and the influence points between adjacent grid nodes refer to the content between two grid nodes that causes the impact match.

[0043] When obtaining the evaluation results, after obtaining the number of grid nodes, the number of grid nodes is used as a benchmark, and then the color area between any two adjacent grid nodes that is likely to have a negative impact on the judgment is calculated. Here, the color area refers to the area with the same or similar color as the grid node.

[0044] After the number of color areas is counted, the following formula is used for processing: Evaluation result = number of grid nodes * K1 + number of color areas * K2; K1 is a positive number, K2 is a negative number, and the sum of the absolute values ​​of K1 and K2 is equal to 1. At this time, the evaluation result is a numerical value. The number of grid nodes has a positive impact on the numerical value, and the color area has a negative impact on the numerical value.

[0045] In some examples, the specific method of parsing the positioning reference image and obtaining the feature parameter set is as follows: Extracting characteristic objects in the positioning reference image, the characteristic objects include moving objects and non-moving objects; Assign type labels to trait objects; Extract local line segment features of feature objects, including shape and color change direction; Calculate the distance value of adjacent local line segment features; The type label, the local line segment feature of the feature object and the distance value of the adjacent local line segment feature are taken as a feature parameter set.

[0046] This method uses the type label, the local line segment features of the feature object and the distance value features of the adjacent local line segment features to generate a feature parameter set. The local line segment includes less feature content, and the type label and the distance value of the adjacent local line segment features are combined to assist in judgment.

[0047] The color change direction is used as a reference feature to distinguish local line features with similar shapes. The color change direction refers to the direction in which the color on the local line feature becomes lighter or darker. There may be one or more color change directions, and each color change direction has a corresponding starting position point, which is generally located at the edge of the local line feature.

[0048] In some examples, the specific method of using feature parameters to search and obtain the region of interest in the positioning reference image is: Use the type label of the feature object to search in the positioning reference image to obtain the suspected area of ​​interest; The local line segment features of the feature object are used to screen the suspected interest regions to obtain the interest regions.

[0049] This method combines two searches. The first one uses type labels to determine the areas where local line segment features may exist, and the second one uses the local line segment features of feature objects for screening. Compared with the method of directly using the local line segment features of feature objects for screening, this method combines two searches can greatly compress the scope of using the local line segment features of feature objects for screening, and can also effectively shorten the search time.

[0050] The specific method of using the local line segment features of the feature object to screen the suspected area of ​​interest is: S401, determining the grayscale value range of the local line segment feature of the feature object and generating a grayscale extraction interval according to the grayscale value range; S402, extracting the suspected region of interest using the grayscale extraction interval to obtain an extracted image; S403, traversing the extracted image using the local line segment features of the feature object; S404, when there are local line segment features of the feature object in the extracted image, the suspected interest region corresponding to the extracted image is retained, otherwise the suspected interest region corresponding to the extracted image is deleted.

[0051] In step S401 to step S404, the grayscale value range of the local line segment features of the feature object is used to generate a grayscale extraction interval, and then the suspected interest area is extracted using the grayscale extraction interval, that is, the content on the suspected interest area that has no association with the local line segment features of the feature object is deleted. This method can exclude a large range of content included in the suspected interest area, and can effectively compress the amount of data processing during traversal.

[0052] The following content is added when traversing in the extracted image using the local line segment features of the feature object: Sequentially divide the grayscale extraction interval into a plurality of sub-grayscale extraction intervals; The local line segment features and suspected interest regions of the feature objects are extracted using the same sub-grayscale extraction interval; Determine whether there is a local line segment feature of a feature object extracted in the same sub-grayscale extraction interval in the extracted image extracted in the same sub-grayscale extraction interval, and obtain an existence result; Determine whether there are local line segment features of the feature object in the extracted image based on the existence result.

[0053] In the above content, the grayscale extraction interval is further divided into multiple sub-grayscale extraction intervals, and then processed using extraction and comparison methods to obtain existence results. Each existence result includes both existence and non-existence, and also corresponds to a specific position.

[0054] Finally, based on the existence results, it is determined whether there are local line segment features of the feature object in the extracted image. The specific method is to first count the existence results (greater than the set number or the set ratio), and then calculate the concentration of the specific positions. For example, if the specific positions exceeding the set number or the set ratio are concentrated in a certain area, it means that there are local line segment features of the feature object here.

[0055] The present application also provides a scene matching device based on drone visual navigation, comprising: An image processing unit is used to acquire an image in the coverage area and segment the acquired image to obtain a segmented image; An image selection unit, used for selecting a segmented image as a positioning reference image; An image parsing unit, used to parse the positioning reference image and obtain a feature parameter set, wherein the feature parameter set includes a plurality of feature parameters; A retrieval processing unit, used to use the characteristic parameters to search in the positioning reference image to obtain a region of interest, where the number of the regions of interest is multiple; A first position determination unit is used to determine an overlapping portion of multiple interest regions and record it as a suspected position region; The second location determination unit is used to screen the suspected location area using the feature parameter set to obtain the current location area.

[0056] Further, selecting a segmented image as a positioning reference image includes: Perform grayscale processing on the segmented image to obtain a grayscale segmented image; Extracting local detail features in the grayscale segmentation image and evaluating the local detail features to obtain evaluation results; Use the evaluation results to sort the grayscale segmented images; The segmented image corresponding to the first grayscale segmented image in the sequential sequence is selected as the positioning reference image.

[0057] Furthermore, extracting local detail features in the grayscale segmentation image and evaluating the local detail features include: Identify feature objects included in the grayscale segmented image, where the feature objects include moving objects and non-moving objects; Construct a reference grid using moving and non-moving objects; Adjust the reference grid so that the number of edges of the grids in the reference grid are consistent; The number of grid nodes and the influence points between adjacent grid nodes are weighted to obtain the evaluation results.

[0058] Furthermore, the positioning reference image is analyzed and processed to obtain a feature parameter set including: Extracting characteristic objects in the positioning reference image, the characteristic objects include moving objects and non-moving objects; Assign type labels to trait objects; Extract local line segment features of feature objects, including shape and color change direction; Calculate the distance value of adjacent local line segment features; The type label, the local line segment feature of the feature object and the distance value of the adjacent local line segment feature are taken as a feature parameter set.

[0059] Furthermore, the feature parameters are used to retrieve and obtain the region of interest in the positioning reference image, including: Use the type label of the feature object to search in the positioning reference image to obtain the suspected area of ​​interest; The local line segment features of the feature object are used to screen the suspected interest regions to obtain the interest regions.

[0060] Further, using the local line segment features of the feature object to screen the suspected interest area includes: Determine the grayscale value range of the local line segment feature of the feature object and generate a grayscale extraction interval according to the grayscale value range; The suspected region of interest is extracted using the grayscale extraction interval to obtain an extracted image; Use the local line segment features of the feature object to traverse in the extracted image; When there are local line segment features of the feature object in the extracted image, the suspected interest region corresponding to the extracted image is retained, otherwise the suspected interest region corresponding to the extracted image is deleted.

[0061] Furthermore, using the local line segment features of the feature object to traverse in the extracted image also includes: Sequentially divide the grayscale extraction interval into a plurality of sub-grayscale extraction intervals; The local line segment features and suspected interest regions of the feature objects are extracted using the same sub-grayscale extraction interval; Determine whether there is a local line segment feature of a feature object extracted in the same sub-grayscale extraction interval in the extracted image extracted in the same sub-grayscale extraction interval, and obtain an existence result; Determine whether there are local line segment features of the feature object in the extracted image based on the existence result.

[0062] In one example, the unit in any of the above devices can be one or more integrated circuits configured to implement the above methods, such as: one or more application specific integrated circuits (ASICs), or, one or more digital signal processors (DSPs), or, one or more field programmable gate arrays (FPGAs), or a combination of at least two of these integrated circuit forms.

[0063] For another example, when the units in the device can be implemented in the form of a processing element scheduling program, the processing element can be a general-purpose processor, such as a central processing unit (CPU) or other processor that can call a program. For another example, these units can be integrated together and implemented in the form of a system-on-a-chip (SOC).

[0064] Various objects such as various messages / information / equipment / network elements / systems / devices / actions / operations / processes / concepts that may appear in this application are named. It can be understood that these specific names do not constitute a limitation on the relevant objects. The names assigned may change with factors such as scenarios, contexts or usage habits. The understanding of the technical meaning of the technical terms in this application should be mainly determined from the functions and technical effects embodied / executed in the technical scheme.

[0065] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0066] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0067] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0068] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0069] It should also be understood that in various embodiments of the present application, the first, second, etc. are only used to indicate that multiple objects are different. For example, the first time window and the second time window are only used to indicate different time windows. They should not have any impact on the time window itself, and the first, second, etc. mentioned above should not impose any limitations on the embodiments of the present application.

[0070] It should also be understood that in the various embodiments of the present application, unless otherwise specified or there is a logical conflict, the terms and / or descriptions between different embodiments are consistent and can be referenced to each other, and the technical features in different embodiments can be combined to form new embodiments according to their internal logical relationships.

[0071] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a computer-readable storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present application. The aforementioned computer-readable storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0072] The present application also provides a scene matching system based on UAV visual navigation, the system comprising: one or more memories for storing instructions; and One or more processors are used to call and run the instructions from the memory to execute the method as described above.

[0073] The present application also provides a computer program product, which includes instructions. When the instructions are executed, the terminal device and the network device perform operations of the terminal device and the network device corresponding to the above method.

[0074] The present application also provides a chip system, which includes a processor for implementing the functions involved in the above content, such as generating, receiving, sending, or processing the data and / or information involved in the above method.

[0075] The chip system may be composed of chips, or may include chips and other discrete devices.

[0076] The processor mentioned in any of the above places can be a CPU, a microprocessor, an ASIC, or one or more integrated circuits for executing programs for controlling the above-mentioned feedback information transmission method.

[0077] In a possible design, the chip system also includes a memory, which is used to store necessary program instructions and data. The processor and the memory can be decoupled and respectively set on different devices, connected by wire or wireless means to support the chip system to implement various functions in the above embodiments. Alternatively, the processor and the memory can also be coupled on the same device.

[0078] Optionally, the computer instructions are stored in a memory.

[0079] Optionally, the memory is a storage unit within the chip, such as a register, a cache, etc. The memory can also be a storage unit within the terminal located outside the chip, such as a ROM or other types of static storage devices that can store static information and instructions, RAM, etc.

[0080] It can be understood that the memory in the present application can be a volatile memory or a non-volatile memory, or can include both volatile and non-volatile memories.

[0081] The nonvolatile memory may be a ROM, a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), or a flash memory.

[0082] The volatile memory may be a RAM, which is used as an external cache. There are many different types of RAM, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), and direct memory bus RAM.

[0083] The embodiments of this specific implementation method are all preferred embodiments of the present application, and are not intended to limit the protection scope of the present application. Therefore, all equivalent changes made based on the structure, shape, and principle of the present application should be included in the protection scope of the present application.

Claims

1. A scene matching method based on UAV visual navigation, characterized in that: include: Acquire an image within the coverage area and perform segmentation processing on the acquired image to obtain a segmented image; Select a segmented image as a positioning reference image; Analyze the positioning reference image and obtain a feature parameter set, where the feature parameter set includes multiple feature parameters; Using the characteristic parameters to search in the positioning reference image, a region of interest is obtained, and the number of the region of interest is multiple; Determine the overlapping parts of multiple interest regions and record them as suspected location regions; The suspected location area is screened using the feature parameter set to obtain the current location area.

2. The scene matching method based on UAV visual navigation according to claim 1 is characterized in that: Selecting a segmented image as a positioning reference image includes: Perform grayscale processing on the segmented image to obtain a grayscale segmented image; Extracting local detail features in the grayscale segmentation image and evaluating the local detail features to obtain evaluation results; Use the evaluation results to sort the grayscale segmented images; The segmented image corresponding to the first grayscale segmented image in the sequential sequence is selected as the positioning reference image.

3. The scene matching method based on UAV visual navigation according to claim 2 is characterized in that: Extracting local detail features from grayscale segmented images and evaluating local detail features include: Identify feature objects included in the grayscale segmented image, where the feature objects include moving objects and non-moving objects; Construct a reference grid using moving and non-moving objects; Adjust the reference grid so that the number of edges of the grids in the reference grid are consistent; The number of grid nodes and the influence points between adjacent grid nodes are weighted to obtain the evaluation results.

4. The scene matching method based on UAV visual navigation according to any one of claims 1 to 3, characterized in that: The positioning reference image is analyzed and processed to obtain a set of feature parameters including: Extracting characteristic objects in the positioning reference image, the characteristic objects include moving objects and non-moving objects; Assign type labels to trait objects; Extract local line segment features of feature objects, including shape and color change direction; Calculate the distance value of adjacent local line segment features; The type label, the local line segment feature of the feature object and the distance value of the adjacent local line segment feature are taken as a feature parameter set.

5. The scene matching method based on UAV visual navigation according to claim 4 is characterized in that: Using feature parameters to retrieve and obtain the region of interest in the positioning reference image includes: Use the type label of the feature object to search in the positioning reference image to obtain the suspected area of ​​interest; The local line segment features of the feature object are used to screen the suspected interest regions to obtain the interest regions.

6. The scene matching method based on UAV visual navigation according to claim 5 is characterized in that: Using the local line segment features of feature objects to screen suspected areas of interest includes: Determine the grayscale value range of the local line segment feature of the feature object and generate a grayscale extraction interval according to the grayscale value range; The suspected region of interest is extracted using the grayscale extraction interval to obtain an extracted image; Use the local line segment features of the feature object to traverse in the extracted image; When there are local line segment features of the feature object in the extracted image, the suspected interest region corresponding to the extracted image is retained, otherwise the suspected interest region corresponding to the extracted image is deleted.

7. The scene matching method based on UAV visual navigation according to claim 6 is characterized in that: The local line segment features of the feature object are used to traverse the extracted image, including: Sequentially divide the grayscale extraction interval into a plurality of sub-grayscale extraction intervals; The local line segment features and suspected interest regions of the feature objects are extracted using the same sub-grayscale extraction interval; Determine whether there is a local line segment feature of a feature object extracted in the same sub-grayscale extraction interval in the extracted image extracted in the same sub-grayscale extraction interval, and obtain an existence result; Determine whether there are local line segment features of the feature object in the extracted image based on the existence result.

8. A scene matching device based on UAV visual navigation, characterized in that: include: An image processing unit is used to acquire an image in the coverage area and segment the acquired image to obtain a segmented image; An image selection unit, used for selecting a segmented image as a positioning reference image; An image parsing unit, used to parse the positioning reference image and obtain a feature parameter set, wherein the feature parameter set includes a plurality of feature parameters; A retrieval processing unit, used to use the characteristic parameters to search in the positioning reference image to obtain a region of interest, where the number of the regions of interest is multiple; A first position determination unit is used to determine an overlapping portion of multiple interest regions and record it as a suspected position region; The second location determination unit is used to screen the suspected location area using the feature parameter set to obtain the current location area.

9. A scene matching system based on UAV visual navigation, characterized in that: The system comprises: one or more memories for storing instructions; and One or more processors, configured to call and execute the instructions from the memory to perform the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer readable storage medium comprises: The program, when the program is executed by a processor, the method according to any one of claims 1 to 7 is executed.

Images