Scene matching and positioning method and scene matching and positioning system based on high-precision map

Abstract

The invention relates to a scene matching and positioning method and a scene matching and positioning system based on a high-precision map. The method comprises the following steps: searching in a visual field range and determining at least one feature object; communication is carried out in the area range, and a flight unit associated with the feature object is searched; determining a relative position relationship between the feature object and a flight unit, wherein the flight unit is arranged around the feature object; and matching the feature objects acquired by the plurality of flight units by using a matching mode to obtain a standard position object, and calculating a local coordinate based on the coordinate of the standard position object. According to the scene matching and positioning method and the scene matching and positioning system based on the high-precision map, the standard position object is obtained by using a collaborative matching processing mode among a plurality of flight units, then the local coordinate is calculated according to the standard position object, and the accurate local coordinate can be quickly determined by the mode.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a scene matching positioning method and a scene matching positioning system based on a high-precision map. 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] Take scene matching navigation as an example. It can achieve accurate positioning of drones in a wider environment. This technology requires building a three-dimensional map through a series of ground or satellite images before flight. During the flight, the drone matches the real-time captured images with this three-dimensional map to calculate its own position and attitude. This technology is particularly important in environments where GPS signals are unavailable or the accuracy is insufficient.

[0004] For example, during the autonomous movement of a drone swarm, when it is disturbed or cannot use satellite positioning, it is necessary to rely on local methods for positioning. However, relying on single-point positioning at this time may result in positioning deviation. The main reason is that the drone obtains part of the ground features. At this time, using comparison method for positioning has a certain probability of error. Summary of the invention

[0005] The present application provides a scene matching positioning method and a scene matching positioning system based on a high-precision map, which uses a collaborative matching processing method between multiple flight units to obtain a standard position object, and then calculates the local coordinates based on the standard position object. This method can quickly determine the accurate local coordinates.

[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 positioning method based on a high-precision map, comprising: Searching within the field of view and determining at least one characteristic object, where the characteristic object belongs to a ground object within the field of view; Communicate within the area to find a flight unit associated with the feature object, where the number of the flight unit is at least one; Determine the relative position relationship between the feature object and the flight unit, and arrange the flight unit around the feature object; Matching the feature objects acquired by multiple flight units using a matching method to obtain a standard position object; Match the standard location object in the high-precision map to obtain the coordinates of the standard location object; Calculates local coordinates based on the coordinates of a standard location object.

[0007] In a possible implementation manner of the first aspect, after determining the feature object, the step further includes extracting identification features of the feature object and constructing a recognition model using the identification features of the feature object; When the matching method is used to match the feature objects obtained by other flight units, the recognition model is used to match the feature objects obtained by other flight units.

[0008] In a possible implementation manner of the first aspect, extracting identification features of the feature object includes: Use edge extraction method to obtain the contour features of feature objects; Selecting a recognition area around the contour feature of the feature object and obtaining a texture feature of the recognition area, the recognition area is located inside the feature object, and the recognition area includes at least two colors; The texture feature curve of the recognition area is calculated, and when the texture feature curve of the recognition area does not meet the requirements, the position of the recognition area is adjusted.

[0009] In a possible implementation manner of the first aspect, calculating a texture feature curve of the identification area includes: Divide the recognition area into multiple rectangular units, and the area of ​​each rectangular unit is the same; Sequentially calculate the total difference between two adjacent rectangular units and use the obtained total difference to draw a texture feature curve; Among them, grayscale processing and regional fuzzy processing are performed on the recognition area; The regional fuzzy processing includes dividing the pixels in the recognition area into multiple groups, calculating the average pixel value of each group of pixels, and using the average pixel value as the pixel value of the corresponding pixel; When the total difference between two adjacent rectangular units is calculated sequentially, the difference between the corresponding pixel points on the two adjacent rectangular units is calculated and the absolute values ​​of the obtained differences are accumulated to obtain the total difference.

[0010] In a possible implementation manner of the first aspect, when searching for a flight unit associated with a feature object, the type of the feature object is used for association.

[0011] In a possible implementation manner of the first aspect, when the type of the feature object is used for association, the method further includes: Selecting auxiliary reference feature objects in the surrounding environment of the feature object; Using the feature object and the auxiliary reference feature object to construct a recognition grid, each endpoint in the recognition grid has a type feature; Use the identification grid to match the identification grids acquired by other flight units.

[0012] In a possible implementation manner of the first aspect, when two recognition grids are matched, it is required that the number of overlapping endpoints exceeds a set number or a set ratio and the overlapping endpoints have a connection relationship.

[0013] In a second aspect, the present application provides a scene matching positioning device based on a high-precision map, comprising: A search unit, used to search within a field of view and determine at least one characteristic object, where the characteristic object belongs to a ground object within the field of view; A communication unit, used for communicating within the area and searching for a flight unit associated with the characteristic object, wherein the number of the flight unit is at least one; a position relationship determination unit, used to determine the relative position relationship between the feature object and the flight unit, the flight unit being arranged around the feature object; A first matching unit is used to match the feature objects acquired by the multiple flight units using a matching method to obtain a standard position object; A second matching unit is used to match the standard position object in the high-precision map to obtain the coordinates of the standard position object; The coordinate calculation unit is used to calculate the local coordinates based on the coordinates of the standard position object.

[0014] In a third aspect, the present application provides a scene matching positioning system based on a high-precision map, 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 positioning method based on a high-precision map provided in this application.

[0021] Figure 2 This is a schematic diagram of an approximate location provided by this application.

[0022] Figure 3 It is a schematic diagram of a comparison identification grid provided in this application.

[0023] Figure 4 This is a schematic diagram of selecting a recognition area around a contour feature provided by the present application.

[0024] Figure 5 This is a schematic diagram provided by the present application for dividing a recognition area into a plurality of rectangular units.

[0025] Figure 6 It is a schematic diagram of a texture characteristic curve provided in this application. DETAILED DESCRIPTION

[0026] The technical solution in this application is further described in detail below in conjunction with the accompanying drawings.

[0027] This application discloses a scene matching positioning method based on high-precision maps, please refer to Figure 1 In some examples, the scene matching positioning method based on high-precision maps disclosed in this application includes the following steps: S101, searching within a field of view and determining at least one characteristic object, where the characteristic object belongs to a ground object within the field of view; S102, communicating within the area to search for a flight unit associated with the feature object, where the number of the flight unit is at least one; S103, determining the relative position relationship between the feature object and the flight unit, and arranging the flight unit around the feature object; S104, matching the feature objects acquired by the multiple flight units using a matching method to obtain a standard position object; S105, matching the standard location object in the high-precision map to obtain the coordinates of the standard location object; S106, calculating the local coordinates based on the coordinates of the standard position object.

[0028] In general, the scene matching positioning method based on high-precision maps disclosed in the present application is mainly used to determine the local position of a drone swarm. There are two ways to calculate the local position of a drone swarm. The first way is to calculate through satellite signals, which involves remote communication. The second way is to use a high-precision map for positioning. The high-precision map carries location coordinates. By comparing with the content included in the high-precision map, the relative position with the compared content can be determined.

[0029] In order to ensure concealment during the silent flight of the drone swarm, the second method is needed to obtain the local position. For a certain feature object in the high-precision map, a single drone can take the approach of circling to obtain complete information. However, for a drone swarm in the silent flight process, this method is no longer applicable. This is because the circling method of a single drone will disrupt the flight integrity of the drone swarm, and the time taken to circling is uncertain, and there is also a potential risk of separation from the drone swarm.

[0030] At the same time, it is also necessary to consider the continuous mobility of the drone swarm, which causes potential positioning errors in the circling method. These errors can be eliminated in the next positioning process, but new positioning errors will be generated at the same time.

[0031] In step S101 of the present application, a search is first performed within the field of view and at least one feature object is determined. This step is performed by a flying unit (drone) in the drone swarm. The feature object obtained at this time is a ground object within the field of view.

[0032] The field of view here refers to the field of view of the image sensor carried by the flight unit when it is working.

[0033] Then, step S102 is executed, in which communication is performed within the area to search for a flight unit associated with the feature object. The number of the flight unit is at least one, that is, the feature object needs to appear in the field of view of at least two flight units.

[0034] Of course, there is a certain probability of error at this time. This is because when determining the uniqueness of the feature object, the category attribute of the feature object is used first. This is a coarse screening method. For example, the feature object at this time is a small building unit. When only the category attribute is used, it may not be possible to determine that multiple flight units see the same feature object, because there are multiple flight units that see multiple feature objects with the same category attributes.

[0035] In step S103, the relative position relationship between the feature object and the flight unit is determined. At this time, the flight unit needs to be arranged around the feature object because the entire surface image of the feature object needs to be obtained. In some possible implementations, a feature object generally requires at least three flight units, which are arranged circumferentially around the feature object.

[0036] Then, in step S104, the feature objects obtained by the multiple flight units are matched using a matching method to obtain a standard position object. As can be seen from the above content, only the type of the feature object is determined before this, that is, it is impossible to determine whether the feature object obtained by the flight unit is the same.

[0037] Step S104 can be regarded as integrating the characteristic objects obtained by the flight units. For example, for the flight units described above being arranged around characteristic objects, it is impossible to determine whether the flight units are around the same characteristic object, but in this step, it can be determined.

[0038] The integration of the feature object is also the integration of the flight unit, and the standard position object obtained here is called the standard position object, which has a complete outer surface. In step S105, the standard position object needs to be matched in the high-precision map to obtain the coordinates of the standard position object.

[0039] The matching method here is to determine the similarity between two images. This part is the same as the content of matching feature objects, which will be further introduced in the subsequent content.

[0040] Finally, in step S106, the local coordinates are calculated based on the coordinates of the standard position object. The standard position object belongs to the high-precision map and has its own coordinates. The coordinates become absolute coordinates. Calculation is performed between the flight unit and the standard position object to obtain relative coordinates. At this time, using absolute coordinates and relative coordinates, the local coordinates of the flight unit can be obtained, that is, the absolute coordinates of the flight unit.

[0041] The acquisition of high-precision maps is generally carried out through high-altitude reconnaissance and photography.

[0042] In some examples, after determining the feature object, it also includes extracting the identification features of the feature object and using the identification features of the feature object to build a recognition model. When using a matching method to match the feature objects obtained by other flight units, the recognition model is used to match the feature objects obtained by other flight units.

[0043] Combining the content mentioned in the previous article, we can see that there is a clear positional relationship between the flying unit and the feature object, which means that when the relative position of a flying unit with the feature object does not change, the entire surface of the feature object cannot be obtained. At this time, it is necessary to cooperate with other flying units.

[0044] For a flight unit flying at low altitude, its field of view will be limited, which means that the image obtained includes less content, resulting in the inability to use a large-scale comparison method to determine the local position. Therefore, a single-body comparison method is used in this application to determine the local position.

[0045] The specific method is to obtain the surface features of a feature object after determining it, and then compare it in the high-precision map. Combining the content mentioned above, we can see that we first use the category attributes of the feature object to determine the approximate location. At this time, the approximate location obtained may be one or more, such as Figure 2 shown.

[0046] Then we need to get the exact location by comparing the feature objects. The method used in this application is to use contour features and local texture features for comparison. Contour features can exclude feature objects of the same type but different shapes, and local texture features can exclude feature objects with the same or similar contour features but different surfaces.

[0047] If there are still multiple feature objects at this time, you need to exclude them using the following method: Selecting auxiliary reference feature objects in the surrounding environment of the feature object; Using the feature object and the auxiliary reference feature object to construct a recognition grid, each endpoint in the recognition grid has a type feature; Use the identification grid to match the identification grids acquired by other flight units.

[0048] The above method uses auxiliary reference feature objects in the surrounding environment of the feature object to determine the position of the feature object. The specific method is to use the feature object and the auxiliary reference feature object to construct a recognition grid, requiring each endpoint in the recognition grid to have a type feature, that is, the recognition grid has both shape and type features.

[0049] See also Figure 3 ,A single mesh shape in the recognition grid generally uses a triangle. ,The way to compare the recognition grids is to place two recognition grids together and ,move some of their endpoints. After the movement, these endpoints overlap in pairs, and the ,overlapping endpoints have the same type features.

[0050] The movement of the endpoint here is allowed as long as the number of grids in the recognition grid is not increased.

[0051] If the two recognition grids can overlap after being processed in the above manner, the number of feature objects finally obtained can be reduced to one. In this process, it is possible to consider appropriately increasing the number of endpoints in the recognition grid to reduce the number of feature objects finally obtained to one.

[0052] The specific method of extracting the identification features of feature objects is as follows: S201, using edge extraction method to obtain contour features of feature objects; S202, selecting a recognition region around the contour feature of the feature object and obtaining a texture feature of the recognition region, wherein the recognition region is located inside the feature object and includes at least two colors; S203, calculating a texture feature curve of the recognition area, and when the texture feature curve of the recognition area does not meet the requirements, adjusting the position of the recognition area.

[0053] In step S201 to step S203, the contour features of the feature object are first obtained, and then a recognition area is selected around the contour features of the feature object ( Figure 4 As shown) and obtain the texture features of the recognition area, it is required that the recognition area is located inside the feature object and that the recognition area includes at least two colors.

[0054] After obtaining the texture features of the recognition area, the texture feature curve of the recognition area is calculated. When the texture feature curve of the recognition area does not meet the requirements, the position of the recognition area is adjusted. Not meeting the requirements here means that the height of the texture feature curve is less than a set fixed parameter value, which means that at least two colors included in the recognition area are too close.

[0055] For the recognition area on two contour features, it is obtained based on the length ratio of the corresponding edges (belonging to the contour features) according to the length proportional relationship.

[0056] The specific method of calculating the texture feature curve of the recognition area is as follows: S301, dividing the recognition area into a plurality of rectangular units ( Figure 5 As shown), the area of ​​each rectangular unit is the same; S302, sequentially calculating the total difference between two adjacent rectangular units and using the obtained total difference to draw a texture feature curve ( Figure 6 shown); This method uses the difference between adjacent rectangular units as the basis for generating a texture feature curve. Because for an identification area, its color will change due to the influence of actual factors such as the acquisition method and the acquisition distance, but this change is a comprehensive change and will not affect the difference between adjacent rectangular units. That is, for the texture feature curve obtained using this method, when the identification area changes as a whole, the texture feature curve will hardly change.

[0057] For the division of the recognition area, the division direction is the same as the trend of the corresponding contour, or all the recognition areas adopt the same division method, such as horizontal division or vertical division.

[0058] At the same time, the recognition area needs to be grayscale processed and the area is blurred. The specific method of blurring is as follows: The regional fuzzy processing includes dividing the pixels in the recognition area into multiple groups, calculating the average pixel value of each group of pixels, and using the average pixel value as the pixel value of the corresponding pixel; When the total difference between two adjacent rectangular units is calculated sequentially, the difference between the corresponding pixel points on the two adjacent rectangular units is calculated and the absolute values ​​of the obtained differences are accumulated to obtain the total difference.

[0059] This fuzzy processing method can further compress the acquisition error and make the change of the total difference less obvious.

[0060] The method of comparing texture feature curves is to place two corresponding texture feature curves in the same coordinate system, and then make the closed area of ​​the two texture feature curves as small as possible by horizontal and / or vertical movement, and then calculate the difference in the height direction, requiring that the maximum difference cannot be greater than a set value.

[0061] In some examples, when using the type of a feature object for association, the following is added: S401, selecting an auxiliary reference feature object in the surrounding environment of the feature object; S402, constructing a recognition grid using the feature object and the auxiliary reference feature object, wherein each endpoint in the recognition grid has a type feature; S403, using the identification grid to match the identification grids acquired by other flight units.

[0062] The content in step S401 to step S403 is to use the auxiliary reference feature objects in the surrounding environment of the feature object to cooperate in determining the position of the feature object. The specific method is to use the feature object and the auxiliary reference feature objects to construct an identification grid, and then use the identification grid to match the identification grid obtained by other flight units.

[0063] At this time, each endpoint in the recognition grid is required to have a type feature, that is, the recognition grid has both shape and type features. The way to compare recognition grids is to place two recognition grids together, and then move some of the endpoints. After these endpoints are moved, they overlap in pairs, and the overlapping endpoints have the same type features.

[0064] The movement of the endpoint here is allowed as long as the number of grids in the recognition grid is not increased.

[0065] In some possible implementations, when two recognition grids are matched, it is required that the number of overlapping endpoints exceeds a set number or a set ratio and the overlapping endpoints have a connection relationship.

[0066] The present application also provides a scene matching positioning device based on a high-precision map, comprising: A search unit, used to search within a field of view and determine at least one characteristic object, where the characteristic object belongs to a ground object within the field of view; A communication unit, used for communicating within the area and searching for a flight unit associated with the characteristic object, wherein the number of the flight unit is at least one; a position relationship determination unit, used to determine the relative position relationship between the feature object and the flight unit, the flight unit being arranged around the feature object; A matching unit, used for matching the feature objects acquired by the multiple flight units using a matching method to obtain a standard position object; The coordinate calculation unit is used to calculate the local coordinates based on the coordinates of the standard position object.

[0067] Furthermore, after determining the feature object, it also includes extracting identification features of the feature object and using the identification features of the feature object to build a recognition model; When the matching method is used to match the feature objects obtained by other flight units, the recognition model is used to match the feature objects obtained by other flight units.

[0068] Furthermore, extracting identification features of feature objects includes: Use edge extraction method to obtain the contour features of feature objects; Selecting a recognition area around the contour feature of the feature object and obtaining a texture feature of the recognition area, the recognition area is located inside the feature object, and the recognition area includes at least two colors; The texture feature curve of the recognition area is calculated, and when the texture feature curve of the recognition area does not meet the requirements, the position of the recognition area is adjusted.

[0069] Furthermore, calculating the texture feature curve of the recognition area includes: Divide the recognition area into multiple rectangular units, and the area of ​​each rectangular unit is the same; Sequentially calculate the total difference between two adjacent rectangular units and use the obtained total difference to draw a texture feature curve; Among them, grayscale processing and regional fuzzy processing are performed on the recognition area; The regional fuzzy processing includes dividing the pixels in the recognition area into multiple groups, calculating the average pixel value of each group of pixels, and using the average pixel value as the pixel value of the corresponding pixel; When the total difference between two adjacent rectangular units is calculated sequentially, the difference between the corresponding pixel points on the two adjacent rectangular units is calculated and the absolute values ​​of the obtained differences are accumulated to obtain the total difference.

[0070] Furthermore, when searching for a flight unit associated with a feature object, the type of the feature object is used for association.

[0071] Furthermore, when the type of the feature object is used for association, it also includes: Selecting auxiliary reference feature objects in the surrounding environment of the feature object; Using the feature object and the auxiliary reference feature object to construct a recognition grid, each endpoint in the recognition grid has a type feature; Use the identification grid to match the identification grids acquired by other flight units.

[0072] Furthermore, when two recognition grids are matched, it is required that the number of overlapping endpoints exceeds a set number or a set ratio and the overlapping endpoints have a connection relationship.

[0073] 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.

[0074] 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).

[0075] 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.

[0076] 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.

[0077] 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.

[0078] 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.

[0079] 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.

[0080] 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.

[0081] 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.

[0082] 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.

[0083] The present application also provides a scene matching positioning system based on a high-precision map, 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.

[0084] 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.

[0085] 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.

[0086] The chip system may be composed of chips, or may include chips and other discrete devices.

[0087] 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.

[0088] 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.

[0089] Optionally, the computer instructions are stored in a memory.

[0090] 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.

[0091] 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.

[0092] The non-volatile memory may be a ROM, a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), or a flash memory.

[0093] 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.

[0094] 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 positioning method based on a high-precision map, characterized in that: include: Searching within the field of view and determining at least one characteristic object, where the characteristic object belongs to a ground object within the field of view; Communicate within the area to find a flight unit associated with the feature object, where the number of the flight unit is at least one; Determine the relative position relationship between the feature object and the flight unit, and arrange the flight unit around the feature object; Matching the feature objects acquired by multiple flight units using a matching method to obtain a standard position object; Match the standard location object in the high-precision map to obtain the coordinates of the standard location object; Calculates local coordinates based on the coordinates of a standard location object.

2. The scene matching positioning method based on high-precision map according to claim 1 is characterized in that: After determining the feature object, it also includes extracting identification features of the feature object and using the identification features of the feature object to build a recognition model; When the matching method is used to match the feature objects obtained by other flight units, the recognition model is used to match the feature objects obtained by other flight units.

3. The scene matching positioning method based on high-precision map according to claim 2 is characterized in that: The identification features of the extracted feature objects include: Use edge extraction to obtain the contour features of the feature object; Selecting a recognition area around the contour feature of the feature object and obtaining a texture feature of the recognition area, the recognition area is located inside the feature object, and the recognition area includes at least two colors; The texture feature curve of the recognition area is calculated, and when the texture feature curve of the recognition area does not meet the requirements, the position of the recognition area is adjusted.

4. The scene matching positioning method based on high-precision map according to claim 3 is characterized in that: Calculating the texture feature curve of the recognition area includes: Divide the recognition area into multiple rectangular units, and the area of ​​each rectangular unit is the same; Sequentially calculate the total difference between two adjacent rectangular units and use the obtained total difference to draw a texture feature curve; Among them, grayscale processing and regional fuzzy processing are performed on the recognition area; The regional fuzzy processing includes dividing the pixels in the recognition area into multiple groups, calculating the average pixel value of each group of pixels, and using the average pixel value as the pixel value of the corresponding pixel; When the total difference between two adjacent rectangular units is calculated sequentially, the difference between the corresponding pixel points on the two adjacent rectangular units is calculated and the absolute values ​​of the obtained differences are accumulated to obtain the total difference.

5. The scene matching positioning method based on high-precision map according to claim 1 is characterized in that: When looking for a flight unit associated with a feature object, the type of the feature object is used for association.

6. The scene matching positioning method based on high-precision map according to claim 5 is characterized in that: When using the type of a feature object for association, it also includes: Selecting auxiliary reference feature objects in the surrounding environment of the feature object; Using the feature object and the auxiliary reference feature object to construct a recognition grid, each endpoint in the recognition grid has a type feature; Use the identification grid to match the identification grids acquired by other flight units.

7. The scene matching positioning method based on high-precision map according to claim 6 is characterized in that: When two recognition grids are matched, the number of overlapping endpoints is required to exceed the set number or the set ratio and the overlapping endpoints are connected.

8. A scene matching positioning device based on a high-precision map, characterized in that: include: A search unit, used to search within a field of view and determine at least one characteristic object, where the characteristic object belongs to a ground object within the field of view; A communication unit, used for communicating within the area and searching for a flight unit associated with the characteristic object, wherein the number of the flight unit is at least one; a position relationship determination unit, used to determine the relative position relationship between the feature object and the flight unit, the flight unit being arranged around the feature object; A first matching unit is used to match the feature objects acquired by the multiple flight units using a matching method to obtain a standard position object; A second matching unit is used to match the standard position object in the high-precision map to obtain the coordinates of the standard position object; The coordinate calculation unit is used to calculate the local coordinates based on the coordinates of the standard position object.

9. A scene matching positioning system based on high-precision maps, 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