Unmanned aerial vehicle autonomous positioning navigation system and method based on visual SLAM

Abstract

The invention relates to the technical field of computer vision and image processing, and discloses an unmanned aerial vehicle autonomous positioning navigation system and method based on visual SLAM. An unmanned aerial vehicle navigation data acquisition module, an unmanned aerial vehicle navigation data calculation module, an unmanned aerial vehicle analysis module, an unmanned aerial vehicle judgment module and an unmanned aerial vehicle execution module are established; data are collected in the unmanned aerial vehicle navigation data acquisition module, data calculation is carried out in the unmanned aerial vehicle navigation data calculation module, and the unmanned aerial vehicle analysis module analyzes whether the positioning precision meets the requirement or not according to the result of the navigation data calculation module. The unmanned aerial vehicle judgment module judges whether error accumulation is within an acceptance range or not and whether a map needs to be updated or not according to a result of the navigation data calculation module, and finally the unmanned aerial vehicle execution module executes a corresponding instruction according to evaluation results of the unmanned aerial vehicle analysis module and the unmanned aerial vehicle judgment module. The method has the advantages of high positioning precision, no error accumulation and fast map updating.

Description

Technical Field

[0001] The present invention relates to the technical fields of computer vision and image processing, and particularly to an unmanned aerial vehicle (UAV) autonomous positioning and navigation system and method based on visual simultaneous localization and mapping (SLAM). Background Art

[0002] In the field of UAV autonomous positioning and navigation, the technology based on visual SLAM (Simultaneous Localization and Mapping) has become a research hotspot due to its adaptability and high efficiency in complex environments. However, the existing technologies still face many challenges, especially in terms of positioning accuracy and error accumulation, as well as map maintenance and management. These problems seriously restrict the further development and application of visual SLAM technology.

[0003] In terms of positioning accuracy and error accumulation, the initial positioning error is a key issue. Once the position estimation of the UAV at startup is inaccurate, or there are errors in the initialization process of the visual SLAM system, these deviations will gradually accumulate during the subsequent positioning process, resulting in the positioning result deviating more and more from the true position. In addition, the long-term error accumulation cannot be ignored. Even if the initial positioning is accurate, over time and with the continuous flight of the UAV, due to factors such as sensor noise and computational rounding errors, the positioning error will continue to accumulate, gradually reducing the accuracy of the map and ultimately possibly leading to positioning failure or navigation errors.

[0004] In terms of map maintenance and management, large-scale map storage and management are also a challenge. During long-term flight missions, the UAV needs to construct a large-scale environmental map, which poses extremely high requirements for map storage and management. If the map data is too large, it will not only occupy a large amount of storage space but also affect the data reading and writing speed and the operating efficiency of the system. In addition, untimely map update is also a problem. When the environment changes, such as the addition or removal of obstacles, the visual SLAM system needs to update the map in a timely manner. However, if the map update is not timely, it will lead to collision risks or the inability to find the correct path during the UAV navigation process.

[0005] Therefore, how to improve the positioning accuracy, suppress error accumulation, and optimize the map storage and update mechanism are the key problems that need to be solved urgently for the current UAV autonomous positioning and navigation system and method based on visual SLAM. Summary of the Invention

[0006] (1) Technical Problems to be Solved

[0007] In view of the deficiencies of the prior art, the present invention provides an autonomous positioning and navigation system and method for an unmanned aerial vehicle (UAV) based on visual simultaneous localization and mapping (SLAM), which has the advantages of high positioning accuracy, no error accumulation, and fast map update, and solves the problems of low positioning accuracy, much error accumulation, and poor map maintenance and management ability in the prior art.

[0008] (II) Technical Solution

[0009] To achieve the above object, the present invention provides the following technical solution: An autonomous positioning and navigation system for an unmanned aerial vehicle based on visual SLAM, including a UAV navigation data acquisition module, a UAV navigation data calculation module, a UAV analysis module, a UAV judgment module, and a UAV execution module;

[0010] The UAV navigation data acquisition module includes an initial positioning data acquisition unit, an environmental impact positioning data acquisition unit, and a map memory registration data acquisition unit;

[0011] The initial positioning data acquisition unit acquires initial positioning data through a visual sensor and an inertial measurement instrument. The environmental impact positioning data acquisition unit acquires environmental impact positioning data through a visual sensor, a noise sensor, and a lidar. The map memory registration data acquisition unit acquires map memory registration data through a historical map storage management library and map update software. The initial positioning data acquisition unit, the environmental impact positioning data acquisition unit, and the map memory registration data acquisition unit are connected to the UAV navigation data calculation module through a network;

[0012] The UAV navigation data calculation module includes a system positioning accuracy calculation unit, an environmental impact calculation unit, and a map memory registration data calculation unit;

[0013] The system positioning accuracy calculation unit calculates the system positioning accuracy Vz according to the initial positioning data. The environmental impact calculation unit calculates the environmental correction coefficient Ru according to the environmental impact positioning data. The map memory registration data calculation unit calculates the optimal deletion upper limit Sj and the optimal deletion lower limit Lj of the map memory according to the map memory registration data. The system positioning accuracy calculation unit, the environmental impact calculation unit, and the map memory registration data calculation unit are connected to the UAV analysis module and the UAV judgment module through a network.

[0014] Preferably, the initial positioning data acquisition unit numbers the data of the i-th true position coordinate according to the characteristics of the initial positioning data, and the number of the i-th true position coordinate is which respectively represent the x, y, and z coordinate values of the i-th true position coordinate in the three-dimensional space.

[0015] Preferably, the environmental impact positioning data acquisition unit numbers the positions of the i-th matching point in the initial screen and the positions of the i-th matching point after being affected by the environment according to the characteristics of the environmental impact positioning data. The positions of the i-th matching point in the initial screen and the positions of the i-th matching point after being affected by the environment are numbered as h i and g i .

[0016] Preferably, the map memory registration data acquisition unit numbers the remaining available storage space in the system and the memory size occupied by the current map data according to the characteristics of the map memory registration data. The remaining available storage space in the system and the memory size occupied by the current map data are numbered as K 1 and K 2 .

[0017] Preferably, the system positioning accuracy calculation unit calculates the system positioning accuracy Vz according to the initial positioning data. The calculation formula is:

[0018]

[0019] In the formula, Vz represents the system positioning accuracy, represents the i-th true position coordinate, represents the i-th system predicted position coordinate, and n represents the total number of position coordinates.

[0020] Preferably, the environmental impact calculation unit calculates the environmental correction coefficient Ru according to the environmental impact positioning data. The calculation formula is:

[0021]

[0022] In the formula, Ru represents the environmental correction coefficient, h i represents the position of the i-th matching point in the initial screen, g i represents the position of the i-th matching point after being affected by the environment, T represents the pose transformation matrix, ||·|| represents the Euclidean distance, represents the pose transformation matrix, and n represents the number of matching points.

[0023] Preferably, the map memory registration data calculation unit calculates the optimal deletion upper limit Sj of the map memory according to the map memory registration data. The calculation formula is:

[0024]

[0025] In the formula, Sj represents the optimal deletion upper limit of the map memory, K 1 represents the remaining available storage space in the system, K 2It represents the memory size occupied by the current map data, and γ represents the preset deletion upper limit ratio of the system.

[0026] Preferably, the map memory registration data calculation unit calculates the optimal deletion lower limit Lj of the map memory according to the map memory registration data, and its calculation formula is:

[0027] Lj = max(K 2 -K 1 , K 2 *β)

[0028] In the formula, Lj represents the optimal deletion lower limit of the map memory, K 1 represents the remaining available storage space in the system, K 2 represents the memory size occupied by the current map data, and β represents the preset deletion lower limit ratio of the system.

[0029] Preferably, the UAV analysis module analyzes whether the positioning accuracy meets the requirements according to the result of the navigation data calculation module;

[0030] The UAV judgment module judges whether the error accumulation is within the acceptable range and whether the map needs to be updated according to the result of the navigation data calculation module;

[0031] The UAV execution module executes corresponding instructions according to the evaluation results of the UAV analysis module and the UAV judgment module.

[0032] Preferably, the UAV autonomous positioning and navigation method based on visual SLAM includes the following steps:

[0033] Step 1: Establish a UAV navigation data acquisition module, a UAV navigation data calculation module, a UAV analysis module, a UAV judgment module and a UAV execution module;

[0034] Step 2: The UAV navigation data acquisition module collects data;

[0035] Step 3: The UAV navigation data calculation module performs data calculation;

[0036] Step 4: The UAV analysis module analyzes whether the positioning accuracy meets the requirements according to the result of the navigation data calculation module;

[0037] Step 5: The UAV judgment module judges whether the error accumulation is within the acceptable range and whether the map needs to be updated according to the result of the navigation data calculation module;

[0038] Step 6: The UAV execution module executes corresponding instructions according to the evaluation results of the UAV analysis module and the UAV judgment module.

[0039] Compared with the prior art, the present invention provides an autonomous positioning and navigation system and method for an unmanned aerial vehicle (UAV) based on visual simultaneous localization and mapping (SLAM), having the following beneficial effects:

[0040] 1. By calculating the system positioning accuracy Vz, the present invention evaluates the reliability of the system used, provides accurate positioning guidance for the UAV, helps improve the accuracy of UAV navigation, can timely detect problems in the positioning process, and provides timely feedback for subsequent system adjustment.

[0041] 2. By calculating the environmental correction coefficient Ru, the present invention can effectively compensate for environmental errors, improve the positioning accuracy, and adjust the environmental correction coefficient Ru according to real-time environmental changes, enabling the system to adapt to different working environments.

[0042] 3. By calculating the optimal deletion upper limit Sj and the optimal deletion lower limit Lj of the map memory through the map memory registration data calculation unit, the present invention can ensure that the map data reasonably occupies the system memory on the premise of meeting the navigation requirements, prevent memory overflow, and ensure the basic functions of the map are not affected by setting the minimum retention amount, maintaining the normal operation and navigation function of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 It is a flowchart of the system of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0044] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0045] Please refer to Figure 1 The flowchart of the system of the present invention. The autonomous positioning and navigation system for an unmanned aerial vehicle based on visual SLAM includes a UAV navigation data acquisition module, a UAV navigation data calculation module, a UAV analysis module, a UAV judgment module, and a UAV execution module;

[0046] The UAV navigation data acquisition module includes an initial positioning data acquisition unit, an environmental impact positioning data acquisition unit, and a map memory registration data acquisition unit;

[0047] The initial positioning data acquisition unit collects initial positioning data through visual sensors and inertial measurement instruments. The environmental impact positioning data acquisition unit collects environmental impact positioning data through visual sensors, noise sensors, and lidar. The map memory registration data acquisition unit collects map memory registration data through the historical map storage management library and map update software. The initial positioning data acquisition unit, the environmental impact positioning data acquisition unit, and the map memory registration data acquisition unit are connected to the UAV navigation data calculation module through the network;

[0048] The UAV navigation data calculation module includes a system positioning accuracy calculation unit, an environmental impact calculation unit, and a map memory registration data calculation unit;

[0049] The system positioning accuracy calculation unit calculates the system positioning accuracy Vz based on the initial positioning data. The environmental impact calculation unit calculates the environmental correction coefficient Ru based on the environmental impact positioning data. The map memory registration data calculation unit calculates the optimal deletion upper limit Sj and the optimal deletion lower limit Lj of the map memory based on the map memory registration data. The system positioning accuracy calculation unit, the environmental impact calculation unit, and the map memory registration data calculation unit are connected to the UAV analysis module and the UAV judgment module through the network.

[0050] The initial positioning data acquisition unit combines visual sensors and inertial measurement instruments to obtain high-precision initial positioning data, providing an accurate starting point for subsequent navigation. At the same time, it numbers the data of the i-th true position coordinate according to the characteristics of the initial positioning data. The number of the i-th true position coordinate is which respectively represent the x, y, and z coordinate values of the i-th true position coordinate in three-dimensional space. This number is used to uniquely identify each true position coordinate in the system for subsequent data processing and navigation calculations.

[0051] The environmental impact positioning data acquisition unit uses visual sensors, noise sensors, and lidar to collect environmental information in real time, enabling it to promptly capture the impact of environmental changes on positioning. At the same time, it numbers the position of the i-th matching point in the initial image and the position of the i-th matching point after being affected by the environment according to the characteristics of the environmental impact positioning data. The position of the i-th matching point in the initial image and the position of the i-th matching point after being affected by the environment are numbered as h i and g i .

[0052] The map memory registration data acquisition unit realizes the efficient management and utilization of map data through the historical map storage management library and map update software, ensuring the accuracy and timeliness of map information. At the same time, according to the characteristics of the map memory registration data, data numbers are assigned to the remaining available storage space in the system and the memory size occupied by the current map data. The remaining available storage space in the system and the memory size occupied by the current map data are numbered as K 1 and K 2 .

[0053] The system positioning accuracy calculation unit calculates the system positioning accuracy Vz based on the initial positioning data, can timely detect positioning deviations, and provides a basis for subsequent adjustments. Its calculation formula is:

[0054]

[0055] In the formula, Vz represents the system positioning accuracy, represents the i-th true position coordinate, represents the i-th system predicted position coordinate, and n represents the total number of position coordinates.

[0056] The advantages are: By calculating the system positioning accuracy Vz, the reliability of system use is evaluated, providing accurate positioning guidance for the UAV, helping to improve the accuracy of UAV navigation, being able to timely detect problems in the positioning process, and providing timely feedback for subsequent system adjustments.

[0057] The environmental impact calculation unit calculates the environmental correction coefficient Ru based on the environmental impact positioning data, dynamically calculates the correction coefficient according to the actual situation, enables the system to adapt to different environmental conditions, and enhances the adaptability of the system. Its calculation formula is:

[0058]

[0059] In the formula, Ru represents the environmental correction coefficient, h i represents the position of the i-th matching point in the initial image, g i represents the position of the i-th matching point after being affected by the environment, T represents the pose transformation matrix, ||·|| represents the Euclidean distance, represents the pose transformation matrix, and n represents the number of matching points.

[0060] The advantages are: By calculating the environmental correction coefficient Ru, environmental errors can be effectively compensated, the positioning accuracy can be improved, and the environmental correction coefficient Ru is adjusted according to real-time environmental changes, enabling the system to adapt to different working environments.

[0061] The map memory registration data calculation unit calculates the optimal deletion upper limit Sj of the map memory according to the map memory registration data. Its calculation formula is:

[0062]

[0063] In the formula, Sj represents the optimal deletion upper limit of the map memory, and K 1 represents the remaining available storage space in the system, and K 2 represents the memory size occupied by the current map data. γ represents the preset deletion upper limit ratio of the system (for example, 0.2 means at most 20% of the current map memory is deleted). The optimal deletion upper limit Sj of the map memory can ensure that when deleting map data, it will not overly occupy the system's available storage space, and at the same time avoid reducing the map accuracy due to deleting too much data.

[0064] The map memory registration data calculation unit calculates the optimal deletion lower limit Lj of the map memory according to the map memory registration data. The calculation formula is:

[0065] Lj = max(K 2 - K 1 , K 2 * γ)

[0066] In the formula, Lj represents the optimal deletion lower limit of the map memory, that is, the minimum retention amount of the map data, ensuring that the basic functions of the map are not affected. K 1 represents the remaining available storage space in the system, and K 2 represents the memory size occupied by the current map data. β represents the preset deletion lower limit ratio of the system (for example, 0.05 means at least 5% of the current map memory is retained). The optimal deletion lower limit Lj of the map memory can ensure that enough map information is retained when deleting data to maintain the normal operation of the system and the navigation function.

[0067] The advantages are: by calculating the optimal deletion upper limit Sj and the optimal deletion lower limit Lj of the map memory through the map memory registration data calculation unit, it can ensure that the map data reasonably occupies the system memory on the premise of meeting the navigation requirements, prevent memory overflow, and ensure that the basic functions of the map are not affected by setting the minimum retention amount, maintaining the normal operation of the system and the navigation function.

[0068] The UAV analysis module analyzes whether the positioning accuracy meets the requirements according to the result of the navigation data calculation module, which helps to timely analyze whether the positioning accuracy meets the requirements, can quickly discover the problems existing in the system, and provides a basis for decision-making;

[0069] The UAV judgment module judges whether the error accumulation is within the acceptable range and whether the map needs to be updated according to the result of the navigation data calculation module. It can effectively judge whether the error accumulation is within the acceptable range, ensure that the positioning accuracy is within a reasonable range, thereby improving the reliability of navigation. At the same time, it judges whether the map needs to be updated according to the positioning situation to ensure the currency and accuracy of the map data;

[0070] The UAV execution module executes corresponding instructions according to the evaluation results of the UAV analysis module and the UAV judgment module;

[0071] The advantages are as follows: when the positioning accuracy is insufficient, the UAV execution module can automatically adjust the flight attitude or perform repositioning; when the map needs to be updated, the module can automatically start the map update program and control the UAV to fly according to the optimized path to avoid collisions and achieve efficient navigation. This module automatically executes corresponding instructions according to the analysis and judgment results to realize the intelligent control of the system, improve the navigation efficiency, and can make corresponding adjustments for different situations, such as adjusting the flight attitude, repositioning or starting the map update program, enhancing the flexibility of the system and the adaptability to environmental changes.

[0072] A method for autonomous positioning and navigation of an unmanned aerial vehicle (UAV) based on visual simultaneous localization and mapping (SLAM) includes the following steps:

[0073] Step 1: Establish a UAV navigation data acquisition module, a UAV navigation data calculation module, a UAV analysis module, a UAV judgment module, and a UAV execution module;

[0074] Step 2: The UAV navigation data acquisition module collects data;

[0075] Step 3: The UAV navigation data calculation module performs data calculation;

[0076] Step 4: The UAV analysis module analyzes whether the positioning accuracy meets the requirements according to the results of the navigation data calculation module;

[0077] Step 5: The UAV judgment module judges whether the error accumulation is within the acceptable range and whether the map needs to be updated according to the results of the navigation data calculation module;

[0078] Step 6: The UAV execution module executes corresponding instructions according to the evaluation results of the UAV analysis module and the UAV judgment module.

[0079] Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art can understand that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. The autonomous positioning and navigation system of UAV based on visual SLAM is characterized by: It includes a UAV navigation data acquisition module, a UAV navigation data calculation module, a UAV analysis module, a UAV judgment module and a UAV execution module; The drone navigation data acquisition module includes an initial positioning data acquisition unit, an environmental impact positioning data acquisition unit, and a map memory registration data acquisition unit; The initial positioning data acquisition unit acquires initial positioning data through visual sensors and inertial measurement instruments, the environmental impact positioning data acquisition unit acquires environmental impact positioning data through visual sensors, noise sensors and laser radars, and the map memory registration data acquisition unit acquires map memory registration data through historical map storage management libraries and map update software. The initial positioning data acquisition unit, the environmental impact positioning data acquisition unit and the map memory registration data acquisition unit are connected to the UAV navigation data calculation module through a network; The drone navigation data calculation module includes a system positioning accuracy calculation unit, an environmental impact calculation unit and a map memory registration data calculation unit; The system positioning accuracy calculation unit calculates the system positioning accuracy Vz based on the initial positioning data, the environmental impact calculation unit calculates the environmental correction coefficient Ru based on the environmental impact positioning data, and the map memory registration data calculation unit calculates the map memory optimal deletion upper limit Sj and the optimal deletion lower limit Lj based on the map memory registration data. The system positioning accuracy calculation unit, the environmental impact calculation unit and the map memory registration data calculation unit are connected to the drone analysis module and the drone judgment module through a network.

2. The UAV autonomous positioning and navigation system based on visual SLAM according to claim 1, characterized in that: The initial positioning data acquisition unit performs data numbering on the i-th real position coordinate according to the initial positioning data characteristics, and the i-th real position coordinate number is They represent the x, y, and z coordinate values ​​of the i-th real position in three-dimensional space respectively.

3. The unmanned aerial vehicle autonomous positioning and navigation system based on visual SLAM according to claim 1, characterized in that: The environmental impact positioning data acquisition unit performs data numbering on the position of the i-th matching point in the initial picture and the position of the i-th matching point after being affected by the environment according to the environmental impact positioning data feature, and the position of the i-th matching point in the initial picture and the position of the i-th matching point after being affected by the environment are numbered h i and g i .

4. The UAV autonomous positioning and navigation system based on visual SLAM according to claim 1, characterized in that: The map memory registration data acquisition unit numbers the remaining available storage space in the system and the memory size occupied by the current map data according to the map memory registration data characteristics, and the remaining available storage space in the system and the memory size occupied by the current map data are numbered K1 and K2 respectively.

5. The UAV autonomous positioning and navigation system based on visual SLAM according to claim 2, characterized in that: The system positioning accuracy calculation unit calculates the system positioning accuracy Vz according to the initial positioning data, and the calculation formula is: In the formula, Vz represents the system positioning accuracy, represents the i-th real position coordinate, represents the i-th system predicted position coordinate, and n represents the total number of position coordinates.

6. The unmanned aerial vehicle autonomous positioning and navigation system based on visual SLAM according to claim 3, characterized in that: The environmental impact calculation unit calculates the environmental correction coefficient Ru according to the environmental impact positioning data, and the calculation formula is: In the formula, Ru represents the environmental correction coefficient, h i represents the position of the i-th matching point in the initial image, g i represents the position of the i-th matching point after being affected by the environment, T represents the pose transformation matrix, ||·|| represents the Euclidean distance, represents the pose transformation matrix, and n represents the number of matching points.

7. The UAV autonomous positioning and navigation system based on visual SLAM according to claim 4, characterized in that: The map memory registration data calculation unit calculates the map memory optimal deletion upper limit Sj according to the map memory registration data, and the calculation formula is: In the formula, Sj represents the optimal upper limit of map memory reduction, K1 represents the remaining available storage space in the system, K2 represents the memory size occupied by the current map data, and γ represents the upper limit ratio of the reduction preset by the system.

8. The UAV autonomous positioning and navigation system based on visual SLAM according to claim 4, characterized in that: The map memory registration data calculation unit calculates the map memory optimal deletion lower limit Lj according to the map memory registration data, and the calculation formula is: Lj=max(K2-K1,K2*β) In the formula, Lj represents the optimal lower limit of map memory deletion, K1 represents the remaining available storage space in the system, K2 represents the memory size occupied by the current map data, and β represents the system preset lower limit deletion ratio.

9. The UAV autonomous positioning and navigation system based on visual SLAM according to claim 8, characterized in that: The drone analysis module analyzes whether the positioning accuracy meets the requirements based on the results of the navigation data calculation module; The drone judgment module judges whether the error accumulation is within an acceptable range and whether the map needs to be updated based on the result of the navigation data calculation module; The drone execution module executes corresponding instructions according to the evaluation results of the drone analysis module and the drone judgment module.

10. The autonomous positioning and navigation method of unmanned aerial vehicle based on visual SLAM is characterized by: The following steps are involved: Step 1: Establish a UAV navigation data acquisition module, a UAV navigation data calculation module, a UAV analysis module, a UAV judgment module and a UAV execution module; Step 2: The UAV navigation data acquisition module collects data; Step 3: The UAV navigation data calculation module performs data calculation; Step 4: The UAV analysis module analyzes whether the positioning accuracy meets the requirements based on the results of the navigation data calculation module; Step 5: The UAV judgment module determines whether the error accumulation is within the acceptable range and whether the map needs to be updated based on the results of the navigation data calculation module; Step 6: The drone execution module executes the corresponding instructions according to the evaluation results of the drone analysis module and the drone judgment module.

Images