Low-altitude panoramic ai detection and identification integrated system

Abstract

The invention discloses a low-altitude panoramic ai detection and recognition integrated system, and relates to the technical field of ai detection and recognition, the system comprises a connecting piece, a bottom plate and a cover plate, the bottom plate is fixedly installed on the bottom surface of the connecting piece and is locked and fixed by seven M5 bolts, the cover plate is fixedly installed on the top surface of the connecting piece, a mainboard module is installed on the top surface of the bottom plate, and the mainboard module is fixedly installed on the bottom surface of the connecting piece. A camera module is arranged on the outer side of the main board module, a support is fixedly connected between the camera module and the bottom plate, fourteen camera units are distributed in a circumferential array mode, so that surrounding environment pictures can be captured in a 360-degree all-around mode, detection and sensing dead angles are avoided, and the detection and sensing efficiency is improved. According to the method, comprehensive environment data is provided for low-altitude flight equipment, the accuracy of an analysis result obtained by subsequent AI data analysis is improved, more accurate data support is provided for various application scenes, and more accurate decision output is provided.

Description

Technical Field

[0001] The present invention relates to the field of AI detection technology, and in particular to a low-altitude panoramic AI detection integrated system. Background Art

[0002] Bird repelling at airports is a key safety measure aimed at preventing birds from colliding with aircraft, thereby ensuring the safe operation of flights. Since birds may collide with aircraft during takeoff, landing or flight, such collisions may not only cause harm to birds, but more importantly, may cause serious damage to aircraft and even cause aviation accidents. Low-altitude panoramic AI exploration refers to the use of artificial intelligence technology to perform panoramic detection and identification of the surrounding environment through drones and other equipment flying at low altitudes. This technology can be applied to various fields, such as agriculture, urban planning, environmental monitoring, etc. It can quickly obtain a wide range of information, and conduct data analysis and processing to provide support for decision-making.

[0003] However, in the existing technology, low-altitude flying equipment such as drones currently need to use cameras to detect the surrounding environment. However, the number of cameras carried by low-altitude flying equipment is limited, and there are certain blind spots in capturing the surrounding environment, making it difficult to capture it comprehensively. As a result, the analysis results obtained by subsequent data analysis are less accurate, and decision-making errors are prone to occur. It is impossible to accurately identify the types of bird flocks and track and identify their flight trajectories, making it difficult to ensure the safety of low-altitude airspace. Summary of the invention

[0004] The purpose of the present invention is to solve the problem that the number of cameras carried by low-altitude flying equipment in the prior art is limited, and there are certain blind spots in capturing the surrounding environment, making it difficult to capture it comprehensively, resulting in low accuracy of the analysis results obtained by subsequent data analysis, prone to decision-making errors, unable to accurately identify the types of bird flocks and track and identify their flight trajectories, and difficult to ensure the safety of low-altitude airspace. A low-altitude panoramic AI exploration integrated system is proposed.

[0005] In order to achieve the above-mentioned objectives, the present invention adopts the following technical solutions: a low-altitude panoramic AI exploration integrated system, comprising a connecting member, a base plate and a cover plate, wherein the base plate is fixedly mounted on the bottom surface of the connecting member and is fastened with seven M5 bolts, the cover plate is fixedly mounted on the top surface of the connecting member, a mainboard module is mounted on the top surface of the base plate, a camera module is arranged on the outside of the mainboard module, a bracket is fixedly connected between the camera module and the base plate, the camera module comprises fourteen camera units, and the fourteen camera units are distributed in a circular array on the inner side of the connecting member, the mainboard module comprises eight mainboard units, wherein the first mainboard unit is electrically connected to the other seven mainboard units, and every two camera units are electrically connected to any one of the other seven mainboard units.

[0006] Preferably, the camera module includes camera unit one, camera unit two, camera unit three, camera unit four, camera unit five, camera unit six, camera unit seven, camera unit eight, camera unit nine, camera unit ten, camera unit eleven, camera unit twelve, camera unit thirteen and camera unit fourteen.

[0007] Preferably, the mainboard module includes mainboard unit 1, mainboard unit 2, mainboard unit 3, mainboard unit 4, mainboard unit 5, mainboard unit 6, mainboard unit 7 and mainboard unit 8, and the mainboard unit 8 is electrically connected to the mainboard unit 1, mainboard unit 2, mainboard unit 3, mainboard unit 4, mainboard unit 5, mainboard unit 6 and mainboard unit 7 respectively through a network cable.

[0008] Preferably, the bracket includes camera bracket 1 and camera bracket 2, and camera bracket 1 is installed on one side of the camera unit.

[0009] Preferably, the camera bracket 2 is installed on the other side of the camera unit, and the bottom ends of the camera bracket 1 and the camera bracket 2 are fixedly connected to the top surface of the base plate.

[0010] Compared with the prior art, the advantages and positive effects of the present invention are:

[0011] 1. In the present invention, by distributing fourteen camera units in a circular array, the present invention can capture the surrounding environment in 360°, avoid the occurrence of blind spots in detection and perception, provide comprehensive environmental data for low-altitude flight equipment, improve the accuracy of analysis results obtained by subsequent AI data analysis, provide more accurate data support for various application scenarios, and provide more accurate decision outputs.

[0012] 2. In the present invention, by connecting every two camera units to a mainboard unit, the mainboard unit connected to the camera unit is connected by signal connection through the mainboard unit eight, and through the function of the network switch of the mainboard unit eight, one of the network cable interfaces is used for external communication, so that the camera unit is connected to the low-altitude flight equipment, providing stable transmission of camera data, thereby facilitating real-time data acquisition of the fourteen camera units and improving the accuracy of low-altitude panoramic AI exploration.

[0013] 3. In the present invention, the types, numbers, stays, entries and exits, and habitat changes of birds active in the airport are monitored in real time to find out the habitat, reproduction, and migration patterns of birds in the airport, reduce the pressure on personnel inspections, and realize real-time monitoring and data analysis of birds in the airport. This has a basis to follow and improves the airport's bird strike prevention and control capabilities. In actual situations, the function of the tracking imaging system is brought into play in the security area to identify bird species that cause high-risk activities, thereby improving the pertinence and effectiveness of bird prevention and greatly improving the scientific and intelligent level of bird strike prevention work. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 The present invention proposes a schematic diagram of the internal structure of a low-altitude panoramic AI exploration integrated system;

[0015] Figure 2 A side view structural diagram of a low-altitude panoramic AI detection integrated system is proposed for the present invention;

[0016] Figure 3 A circuit connection diagram of a low-altitude panoramic AI exploration integrated system is proposed for the present invention;

[0017] Figure 4 A schematic diagram of a power supply method for a low-altitude panoramic AI exploration integrated system is proposed for the present invention;

[0018] Figure 5 A system block diagram of a low-altitude panoramic AI exploration integrated system is proposed for the present invention;

[0019] Figure 6 A three-difference frame method flow chart of a low-altitude panoramic AI detection integrated system is proposed for the present invention;

[0020] Figure 7 The present invention proposes a device block diagram of a low-altitude panoramic AI exploration integrated system.

[0021] Legend: 1. Camera unit one; 2. Camera unit two; 3. Camera unit three; 4. Camera unit four; 5. Camera unit five; 6. Camera unit six; 7. Camera unit seven; 8. Camera unit eight; 9. Camera unit nine; 10. Camera unit ten; 11. Camera unit eleven; 12. Camera unit twelve; 13. Camera unit thirteen; 14. Camera unit fourteen; 15. Mainboard unit one; 16. Mainboard unit two; 17. Mainboard unit three; 18. Mainboard unit four; 19. Mainboard unit five; 20. Mainboard unit six; 21. Mainboard unit seven; 22. Mainboard unit eight; 23. Camera bracket one; 24. Camera bracket two; 25. Connector; 26. Bottom plate; 27. Cover plate. DETAILED DESCRIPTION

[0022] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described below in conjunction with the accompanying drawings and embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0023] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways than those described herein. Therefore, the present invention is not limited to the specific embodiments of the following disclosure.

[0024] Embodiment 1

[0025] like Figure 1-4 As shown, the present invention provides an integrated low-altitude panoramic AI exploration system, including a connecting member 25, a base plate 26 and a cover plate 27. The base plate 26 is fixedly installed on the bottom surface of the connecting member 25, and the cover plate 27 is fixedly installed on the top surface of the connecting member 25. A mainboard module is installed on the top surface of the base plate 26, and a camera module is arranged on the outside of the mainboard module. A bracket is fixedly connected between the camera module and the base plate 26. The camera module includes fourteen camera units, and the fourteen camera units are distributed in a circular array on the inner side of the connecting member 25. The mainboard module includes eight mainboard units, wherein the first mainboard unit is electrically connected to the other seven mainboard units, and every two camera units are electrically connected to any one of the other seven mainboard units, and are connected using a CSI-2 protocol interface.

[0026] The specific settings and functions of this embodiment are described in detail below. By distributing fourteen camera units in a circular array, the present invention can capture the surrounding environment in 360°, avoid the occurrence of blind spots in detection and perception, provide comprehensive environmental data for low-altitude flight equipment, improve the accuracy of analysis results obtained by subsequent AI data analysis, provide more accurate data support for various application scenarios, and provide more accurate decision outputs. Through the arrangement of the connector 25, the base plate 26 and the cover plate 27, the internal mainboard module and the camera module are protected to avoid damage.

[0027] Embodiment 2

[0028] like Figure 1-4As shown, the camera module includes camera unit 1, camera unit 2, camera unit 3, camera unit 4, camera unit 5, camera unit 6, camera unit 7, camera unit 8, camera unit 9, camera unit 10, camera unit 11, camera unit 12, camera unit 13, and camera unit 14, and the mainboard module includes mainboard unit 15, mainboard unit 2, mainboard unit 3, mainboard unit 4, mainboard unit 5, mainboard unit 6, mainboard unit 7, and mainboard unit 8, and the mainboard unit 8 is connected to the mainboard unit 15, the mainboard unit 2, 16, The mainboard unit three 17, the mainboard unit four 18, the mainboard unit five 19, the mainboard unit six 20 and the mainboard unit seven 21 are electrically connected through a network cable. The mainboard unit eight 22 has a total of eight network cable interfaces and has the function of a network switch. One of the network cable interfaces is used for external communication and is led to the outside together with the power interface. The eight mainboard units are powered by the same power supply, and all power supplies are connected together. The bracket includes a camera bracket one 23 and a camera bracket two 24. The camera bracket one 23 is installed on one side of the camera unit, and the camera bracket two 24 is installed on the other side of the camera unit. The bottom ends of the camera bracket one 23 and the camera bracket two 24 are fixedly connected to the top surface of the base plate 26.

[0029] The effect achieved by the entire embodiment is that by connecting every two camera units to a mainboard unit, the mainboard unit connected to the camera unit is signal-connected through the mainboard unit 822, and through the function of the mainboard unit 822 network switch, one of the network cable interfaces is used for external communication, so that the camera unit is connected to the low-altitude flight equipment, providing stable transmission of camera data, thereby facilitating real-time data acquisition of fourteen camera units and improving the accuracy of low-altitude panoramic AI detection.

[0030] Embodiment 3

[0031] like Figure 5-7 As shown, the airport bird recognition technology based on motion target detection of the system of the present invention mainly uses the video surveillance system to capture the dynamic situation around the airport, and analyzes the characteristics of the moving target to achieve the recognition and tracking of the flying bird, which specifically includes: data collection, data preprocessing, feature extraction, target detection and tracking, and decision support;

[0032] Data collection: 360-degree high-resolution cameras are deployed around the airport to ensure that bird movements can be captured from multiple angles. These devices can provide high-quality real-time video streams and motion data to support subsequent data processing and analysis;

[0033] Data preprocessing: The collected data needs to be preprocessed, including steps such as noise removal, image enhancement, and motion detection, to improve the accuracy and efficiency of subsequent processing;

[0034] Feature extraction: Use advanced image processing technology and machine learning algorithms to extract the features of flying birds from pre-processed data. These features include shape, size, flight speed, flight trajectory, etc. After each update of the tracked target, calculate the overall motion path length, speed, direction, distance from the starting point to the current point, etc. of the target as feature data, and select targets based on these features;

[0035] Target detection and tracking: Based on the extracted features, flying birds can be accurately detected, and the code stream of the camera can be used to detect and track the flying birds around the airport in real time. The system can automatically identify and track flying birds, and record their flight paths and behavior patterns;

[0036] Specifically include: AI target detection, moving target tracking and fusion judgment;

[0037] The process of AI target detection is:

[0038] 1. Image acquisition: Image data is acquired frame by frame from the camera video stream.

[0039] 2. Image preprocessing: perform preprocessing operations such as resizing, normalization, and color space conversion on image data.

[0040] 3. Data input: The preprocessed image data is input into the embedded neural network module of the device chip.

[0041] 4. Model reasoning: Run the pre-trained object detection model on the embedded neural network module to perform real-time image analysis and object detection.

[0042] 5. Result processing: After the model inference is completed, the detected target information will be output, including the category, location, confidence level, etc. of each target, and then NMS (non-maximum suppression) processing will be performed to select the optimal result;

[0043] The process of moving target tracking is:

[0044] 1. Moving target detection: The moving target is determined by using the detection principle of strong correlation between adjacent frames of video images.

[0045] Since there are some defects in the two-difference frame method for moving target detection, such as holes in the detection results when the moving objects themselves have similar colors, it is difficult to detect the target when the position changes slowly, and the light changes have a great impact, so the three-difference frame method is used here;

[0046] 2. Target matching: After detecting an active target, it is matched with the previously detected target. After detecting a moving target in each frame, it needs to be matched with the target tracked in the previous frame to ensure the continuity of target detection and avoid duplicate counting.

[0047] The target matching adopts the minimum Euclidean distance matching method, that is, the Euclidean distance matrices of all targets in the current frame and the tracked targets in the previous frame are calculated respectively, and then the Pearson correlation coefficient of the two matrices is calculated. The current frame target with the largest correlation coefficient is selected as the matching target for the tracked target of the previous frame, and the current frame target that does not meet the matching requirements is added as a new tracking target.

[0048] Euclidean distance calculation formula:

[0049]

[0050] Where: d is the Euclidean distance between two points, (x1, y1) and (x2, y2) are the coordinates of the two points.

[0051] The Pearson correlation coefficient, also known as the Pearson product-moment correlation coefficient, is a measure of the correlation between two variables. The Pearson correlation coefficient between two variables is defined as the quotient of the covariance and standard deviation between the two variables:

[0052]

[0053] Among them: r is the Pearson correlation coefficient, X and Y are two groups of variables, is the average value of the two variables X and Y.

[0054] 3. Feature identification: After each update of the tracked target, the overall motion path length, speed, direction, distance from the starting point to the current point, etc. of the target are calculated as feature data, and the target is screened based on these features.

[0055] 4. Result output: Output the bounding box coordinates of the target filtered in the previous step;

[0056] The process of fusion judgment is as follows: since target tracking is a continuous process, even if the deep learning model fails to detect an error, the target can be detected again in the subsequent detection process to make up for it, which can effectively improve the accuracy of bird detection. After obtaining the results of AI target detection and moving target tracking, fusion judgment is performed to ensure the accuracy of the final target detection and avoid repeated counting.

[0057] Decision support: The system can not only identify and track flying birds, but also provide decision support based on the birds’ behavior and flight trajectory. For example, when a flock of birds is detected approaching the flight path, the system can automatically issue an alarm, prompting the airport control center to take measures, such as activating bird repellent equipment.

[0058] In the present invention: by distributing fourteen camera units in a circular array, the present invention can capture the surrounding environment in an all-round manner at 360°, avoiding the occurrence of blind spots in detection and perception, by connecting every two camera units to a mainboard unit, and connecting the mainboard unit connected to the camera unit by signal through the mainboard unit eight 22, and through the function of the mainboard unit eight 22 network switch, one of the network cable interfaces is used for external communication, so that the camera unit is connected to the low-altitude flight equipment, providing stable transmission of multiple groups of camera data, and the airport bird recognition technology based on motion target detection mainly uses a video surveillance system to capture the dynamic situation around the airport, and analyzes the characteristics of the motion target to achieve recognition and tracking of birds.

[0059] The above description is only a preferred embodiment of the present invention and does not limit the present invention in other forms. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes and apply it to other fields. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A low-altitude panoramic AI exploration integrated system, characterized by: The invention comprises a connecting member (25), a bottom plate (26) and a cover plate (27), wherein the bottom plate (26) is fixedly mounted on the bottom surface of the connecting member (25), the cover plate (27) is fixedly mounted on the top surface of the connecting member (25), a mainboard module is mounted on the top surface of the bottom plate (26), a camera module is arranged on the outside of the mainboard module, a bracket is fixedly connected between the camera module and the bottom plate (26), the camera module comprises fourteen camera units, the fourteen camera units are arranged in a circular array on the inside of the connecting member (25), the mainboard module comprises eight mainboard units, the first mainboard unit is electrically connected to the other seven mainboard units, and every two camera units are electrically connected to any one of the other seven mainboard units.

2. According to claim 1, a low-altitude panoramic AI exploration integrated system is characterized by: The camera module includes camera unit one (1), camera unit two (2), camera unit three (3), camera unit four (4), camera unit five (5), camera unit six (6), camera unit seven (7), camera unit eight (8), camera unit nine (9), camera unit ten (10), camera unit eleven (11), camera unit twelve (12), camera unit thirteen (13) and camera unit fourteen (14).

3. The low-altitude panoramic AI detection integrated system according to claim 1 is characterized by: The mainboard module comprises a mainboard unit 1 (15), a mainboard unit 2 (16), a mainboard unit 3 (17), a mainboard unit 4 (18), a mainboard unit 5 (19), a mainboard unit 6 (20), a mainboard unit 7 (21) and a mainboard unit 8 (22), and the mainboard unit 8 (22) is electrically connected to the mainboard unit 1 (15), the mainboard unit 2 (16), the mainboard unit 3 (17), the mainboard unit 4 (18), the mainboard unit 5 (19), the mainboard unit 6 (20) and the mainboard unit 7 (21) respectively through a network cable.

4. The low-altitude panoramic AI detection integrated system according to claim 1 is characterized by: The bracket comprises a camera bracket 1 (23) and a camera bracket 2 (24), wherein the camera bracket 1 (23) is installed on one side of the camera unit.

5. The low-altitude panoramic AI detection integrated system according to claim 4 is characterized by: The second camera bracket (24) is installed on the other side of the camera unit, and the bottom ends of the first camera bracket (23) and the second camera bracket (24) are fixedly connected to the top surface of the bottom plate (26).

Images