Parking space monitoring method and system based on vehicle detection, storage medium and terminal

Abstract

The invention discloses a parking space monitoring method and system based on vehicle detection, a storage medium and a terminal. Obtaining an image through a camera installed in the parking lot to determine the parking position coordinate of the current entering vehicle; comparing the parking position coordinate of the current entering vehicle with each pre-stored parking space coordinate in the parking space set; if the position overlap between the coordinates and the rectangular frames represented by all the coordinates is smaller than a low preset threshold value, determining that the parking position coordinate of the current entering vehicle is a new parking space; if the position overlapping between the rectangular frames represented by the two coordinates reaches a high preset threshold value, determining that the parking position of the current entering vehicle and the comparison pre-stored parking space are repeated parking spaces; according to the invention, the specific position of the vehicle is identified by using the target detection technology, and high accuracy and efficiency can be maintained even if visual information is affected by severe weather.

Description

Technical Field

[0001] The present invention relates to the technical field of parking space detection, and particularly to a parking space monitoring method and system, a storage medium and a terminal based on vehicle detection. Background Art

[0002] With the acceleration of the global urbanization process, the problem of urban parking has become increasingly prominent. The busy urban life requires more efficient and intelligent parking solutions to relieve traffic congestion, improve the utilization efficiency of parking lots, and at the same time provide a more convenient parking experience for drivers. In this context, the image-based parking space recognition technology in the field of autonomous driving has performed very well, and thus the performance of these algorithms under extremely harsh weather conditions deserves more attention because they can provide greater help and effectiveness in these challenging environments.

[0003] Traditional image-based parking space detection methods first involve preprocessing the input image using a series of basic image processing techniques. This usually includes applying filtering techniques to smooth the image and reduce noise. In addition, edge detection algorithms (such as Sobel or Canny algorithms) are used to highlight the edge information in the image, while morphological operations such as dilation and erosion are used to strengthen or refine the structural features in the image. Next, key features are identified through feature extraction, including straight lines, corner points, and specific geometric shapes. Such as the Hough transform, Harris corner detector, or Shi-Tomasi method. And using object detection algorithms such as Faster R-CNN or YOLO for parking space detection is a method that does not need to rely on parking space markings. In this method, the entire image with parking space annotations is used as training data, enabling the model to learn to identify and locate parking spaces and at the same time determine whether they are occupied.

[0004] Although these methods can effectively identify and confirm the status of parking spaces in the parking lot, they face some challenges in practical applications. In extremely harsh weather such as heavy rain, heavy snow, and thick fog, there are often situations where the parking space lines are blocked or unclear. And even for methods that do not rely on parking space markings, when detecting empty parking spaces, due to their high visual consistency with the surrounding environment, the model often has difficulty distinguishing, as well as the noise interference brought by different harsh weather conditions, resulting in inaccurate detection results. Summary of the Invention

[0005] Aiming at the inaccurate detection results in the face of bad weather pointed out in the background art, the purpose of the present invention is to propose a parking space monitoring method and system, a storage medium and a terminal based on vehicle detection.

[0006] To achieve the purpose of the present invention, the technical solutions proposed by the present invention are as follows:

[0007] In the first aspect

[0008] The present application provides a parking space monitoring method based on vehicle detection, including the following steps:

[0009] Step S1: Obtain an image through a camera installed in the parking lot to determine the parking position coordinates of the currently entering vehicle;

[0010] Step S2: Determine whether the parking space set is empty; if it is empty, add the parking position coordinates of the currently entering vehicle, the occupied flag status of the parking space, and the number of times n the parking space has been detected to the parking space set as a pre-stored parking space. If it is not empty, jump to Step S3; where the initial value of n is 1;

[0011] Step S3: Compare the parking position coordinates of the currently entering vehicle with the coordinates of each pre-stored parking space in the parking space set;

[0012] If the position overlap between the rectangular frames represented by all coordinates is less than a low preset threshold, it is determined that the parking position coordinates of the currently entering vehicle are a new parking space, and the parking position coordinates of the currently entering vehicle, the occupied flag status of the parking space, and the number of times n the parking space has been detected are added to the parking space set as a pre-stored parking space; where the initial value of n is 1;

[0013] If the position overlap between the rectangular frames represented by two coordinates reaches a high preset threshold, it is determined that the parking position of the currently entering vehicle and the compared pre-stored parking space are duplicate parking spaces. Correct the parking position coordinates of the currently entering vehicle, and add the corrected coordinates, the occupied flag status of the parking space, and the updated number of times the parking space has been detected to the parking space set as a pre-stored parking space. After that, delete the compared pre-stored parking space;

[0014] When the number of pre-stored parking spaces in the parking space set reaches the preset number of parking spaces and there is no increase within a preset time period, the pre-stored parking space coordinates in the parking space set are the coordinates of all parking spaces;

[0015] Step S4: Based on all the parking spaces obtained in Step S3.

[0016] Wherein, it further includes the step of modifying the occupied flag status of the pre-stored parking space to an unoccupied flag status when it is detected that the vehicle on a certain pre-stored parking space leaves the parking lot.

[0017] Wherein, in Step S1, it includes the step of using a trained vehicle detection model to perform vehicle detection on the image. Among them, the original image data set used for training the vehicle detection model is generated by adding rain, snow, and fog weather effects through an artificial simulation weather effect algorithm to generate a simulated image data set.

[0018] Among them, the method can be automatically called and executed by an automated script tool.

[0019] In a second aspect

[0020] This application provides a parking space monitoring system based on vehicle detection, including the following units: an image acquisition unit, a parking space set judgment unit, a comparison unit, and a parking space status monitoring unit;

[0021] The image acquisition unit is used to determine the parking position coordinates of the currently entering vehicle by acquiring images through a camera installed in the parking lot;

[0022] The parking space set judgment unit is used to judge whether the parking space set is empty; if it is empty, add the parking position coordinates of the currently entering vehicle, the occupied flag status of the parking space, and the number of times n that the parking space has been detected to the parking space set as a pre-stored parking space. If it is not empty, execute the comparison unit; where the initial value of n is 1;

[0023] The comparison unit is used to compare the parking position coordinates of the currently entering vehicle with each pre-stored parking space coordinate in the parking space set;

[0024] If the position overlap between the rectangle frames represented by all coordinates is less than a low preset threshold, it is determined that the parking position coordinates of the currently entering vehicle are a new parking space, and the parking position coordinates of the currently entering vehicle, the occupied flag status of the parking space, and the number of times n that the parking space has been detected are added to the parking space set as a pre-stored parking space; where the initial value of n is 1;

[0025] If the position overlap between the rectangle frames represented by two coordinates reaches a high preset threshold, it is determined that the parking position of the currently entering vehicle and the compared pre-stored parking space are duplicate parking spaces, correct the parking position coordinates of the currently entering vehicle, and add the corrected coordinates, the occupied flag status of the parking space, and the updated number of times the parking space has been detected to the parking space set as a pre-stored parking space. After that, delete the compared pre-stored parking space;

[0026] When the number of pre-stored parking spaces in the parking space set reaches the preset number of parking spaces and there is no increase within the preset time period, the pre-stored parking space coordinates in the parking space set are all the parking space coordinates;

[0027] The parking space status monitoring unit is used to judge the occupancy status of the parking space based on all the parking space coordinates obtained by the comparison unit.

[0028] Among them, it further includes a status modification unit, which is used to modify the occupied flag status of a pre-stored parking space to an unoccupied flag status when it is detected that the vehicle on a certain pre-stored parking space leaves the parking lot.

[0029] Among them, the image acquisition unit is further configured to perform vehicle detection on the image using a trained vehicle detection model. The original image dataset used for training the vehicle detection model is generated by adding rain, snow, and fog weather effects through an artificially simulated weather effect algorithm to generate a simulated image dataset.

[0030] Among them, the system can be automatically called and executed through an automated script tool.

[0031] In a third aspect

[0032] This application provides a storage medium in which at least one instruction, at least one program, a code set, or an instruction set is stored. The at least one instruction, the at least one program, the code set, or the instruction set is loaded and executed by a processor to implement the above-mentioned parking space monitoring method based on vehicle detection.

[0033] In a fourth aspect

[0034] This application provides an electronic terminal. The electronic terminal includes a processor and a memory. At least one instruction, at least one program, a code set, or an instruction set is stored in the memory. The at least one instruction, the at least one program, the code set, or the instruction set is loaded and executed by the processor to implement the above-mentioned parking space monitoring method based on vehicle detection.

[0035] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0036] The solution provided by the present invention uses the images captured by the cameras installed in the parking lot and utilizes object detection technology to identify the specific positions of vehicles, thereby inferring which parking spaces are occupied. In addition, when a vehicle leaves, the corresponding position is updated to the idle state. The dynamic update mechanism of this method can update and record the status of each parking space in real time, including the initially unoccupied idle parking spaces, and can maintain high accuracy and efficiency even when visual information is affected by bad weather. Through this technology, parking resources can be effectively monitored and managed under various weather conditions, greatly improving the applicability and reliability of the parking space detection system.

[0037] In addition, this application utilizes an artificially simulated weather effect algorithm, which can superimpose various meteorological effects, such as rain, snow, and fog, on the selected original image dataset. This technology enables the generation of a set of simulated image datasets, which widely cover various climate conditions. Such a method greatly enriches the diversity of the dataset and helps to improve the accuracy and robustness of the parking space detection system in complex climates.

[0038] In addition, the solution of the present application can be applied to an automated script tool, through which the entire task process of simulating weather image synthesis and parking space detection can be automated. This kind of application improves the efficiency and repeatability of experiments, making the parking space detection under various specific weather conditions more efficient and accurate. Description of the Drawings

[0039] Figure 1 Schematic flow chart of the method provided by the embodiment of the present application;

[0040] Figure 2 Schematic diagram of the intersection of rectangle A and rectangle B in the embodiment of the present application. Detailed Description of the Embodiment

[0041] The present invention will be further described in detail below in conjunction with the drawings and specific embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0042] As Figure 1 shown, the present application provides a parking space monitoring method based on vehicle detection, including the following steps:

[0043] Step S1: Determine the parking position coordinates of the currently entering vehicle by obtaining an image through a camera installed in the parking lot;

[0044] Step S2: Determine whether the parking space set is empty; if it is empty, add the parking position coordinates of the currently entering vehicle, the occupied flag state of the parking space, and the number of times n the parking space is detected to the parking space set as a pre-stored parking space. If it is not empty, jump to step S3; where the initial value of n is 1;

[0045] Step S3: Compare the parking position coordinates of the currently entering vehicle with the coordinates of each pre-stored parking space in the parking space set;

[0046] If the position overlap between the rectangle frames represented by all coordinates is less than the low preset threshold, it is determined that the parking position coordinates of the currently entering vehicle are a new parking space, and the parking position coordinates of the currently entering vehicle, the occupied flag state of the parking space, and the number of times n the parking space is detected are added to the parking space set as a pre-stored parking space; where the initial value of n is 1;

[0047] If the position overlap between the rectangle frames represented by two coordinates reaches the high preset threshold, it is determined that the parking position of the currently entering vehicle and the compared pre-stored parking space are duplicate parking spaces, correct the parking position coordinates of the currently entering vehicle, and add the corrected coordinates, the occupied flag state of the parking space, and the updated number of times the parking space is detected to the parking space set as a pre-stored parking space. After that, delete the compared pre-stored parking space;

[0048] When the number of pre-stored parking spaces in the parking space set reaches the preset number of parking spaces and there is no increase within the preset time period, the pre-stored parking space coordinates in the parking space set are the coordinates of all parking spaces;

[0049] Step S4: Based on all the parking spaces obtained in Step S3.

[0050] Further, it includes the step of modifying the occupied identification status of the pre-stored parking space to the unoccupied identification status when it is detected that the vehicle on a certain pre-stored parking space leaves the parking lot.

[0051] In Step S1, it includes the step of using a trained vehicle detection model to detect vehicles in the image. The original image dataset used for training the vehicle detection model is generated by adding rain, snow, and fog weather effects through an artificially simulated weather effect algorithm to form a simulated image dataset.

[0052] It should be noted that the image restoration process and complex special weather synthesis algorithms are often accompanied by high computational costs and possible information losses. In this case, a fast and simple synthesis algorithm can still perform well and thus is applied. In this application, the artificial rainy weather algorithm simulates the blurring effect generated by the rapid movement of raindrops in the air relative to the camera sensor during the falling process through motion blur technology. The key to this technology lies in the use of a motion blur kernel, and by performing a convolution operation on the original image, the effect of simulating the raindrop movement trajectory is achieved. To simulate the visual effect in a foggy environment, Gaussian blur is adopted as one of the core technologies. Gaussian blur simulates the blurring effect of the edges of distant objects under foggy conditions by smoothing the image. This method is based on the Gaussian function and performs a weighted average on each pixel point of the image, and the weight is determined by the distance of the pixel point to the center of the Gaussian distribution. By adjusting the parameters of Gaussian blur, such as the blur radius, the algorithm can control the intensity of the blurring effect, thereby simulating foggy environments with different densities from light fog to thick fog. For the simulation effect of snow days, image fusion is used to overlay the foggy effect and the snowflake effect in a certain proportion on different layers, aiming to simulate the visual characteristics in a real snow day environment, so as to achieve a more real and comprehensive simulation effect.

[0053] The relevant algorithms adopted in the method of this application are as follows:

[0054]

[0055]

[0056]

[0057] The method can be automatically called and executed through an automated script tool.

[0058] The automated script tool can enhance the performance of the parking space recognition system under diverse and complex environmental conditions by simulating actual scenarios under different climate conditions. The automated script tool can introduce a customized artificial effect algorithm, which makes it possible to superimpose various meteorological effects on the selected original image dataset, including but not limited to rain, snow, fog, etc., thereby creating a set of simulated image datasets covering a wide range of climate conditions. With the successful implementation of this simulated climate effect algorithm, this script then activates multiple weather recognition models for invocation. These models combine deep learning technology to conduct in-depth analysis and recognition of the meteorological conditions in the images, and can accurately identify the specific weather type simulated by the current image. After successfully identifying the specific climate conditions, the script will immediately start the pre-configured deep network model, first perform the parking space recognition task on the processed simulated climate mixed dataset, and then conduct parking space recognition for the scenarios under specific climate conditions.

[0059] It should be noted that referring to Figure 2 , the specific steps of this method are as follows:

[0060] (1) Initialize the parking space set Use the trained vehicle detection model to detect vehicles in the image. Figure 2 In a1 , the rectangular box A represents the parking space occupied by the new vehicle. The upper left coordinates (x a1 , y a2 ) and the lower right coordinates (x a2 , y a1 ) of the rectangular box A, and the vehicle position occupancy information (x a1 , y a2 , x a2 , 1), where 1 represents that there is a vehicle at this position, and 0 is used to represent no vehicle; and the number of times detected (initial count is 1) is added to the set Φ.

[0061] (2) For each detection, compare the detected target with each pre-stored parking space in the set. The rectangular box B is the pre-stored parking space, and the intersection over union (IoU) is used as a measurement index:

[0062]

[0063] Among them, A and B are two matrix boxes. Assume that the target rectangular box A currently detected has a vehicle, and its position occupancy information is (x a1 , y a1 , x a2 , y a2 , 1), and the matrix box B in the set has no vehicle, and its position occupancy information is (xb1 , y b1 , x b2 , y b2 , 0), if there is an intersection between the two, then for the intersection area,

[0064] The upper left corner coordinates are:

[0065] x1 = max(x a1 , x b1 ), y1 = max(y a1 , y b1 ) (1)

[0066] The lower right corner coordinates are:

[0067] x2 = max(x a2 , x b2 ), y2 = max(y a2 , y b2 ) (2)

[0068] From the coordinates of the intersection, the area calculation formula for the intersection can be obtained:

[0069] intersection = max(x2 - x1 + 1.0, 0)max(y2 - y1 + 1.0, 0) (3)

[0070] The area of rectangle A is:

[0071] S A = (x a2 - x a1 + 1.0)(y a2 - y a1 + 1.0) (4)

[0072] The area of rectangle B is:

[0073] S B = (x b2 - x b1 + 1.0)(y b2 - y b1 + 1.0) (5)

[0074] From the areas of A and B and the intersection area, the union area can be obtained:

[0075] union = S A + S B - intersection (6)

[0076] Calculate the intersection over union ratio according to the intersection area formula (3) and the union area formula (6):

[0077]

[0078] When the IoU is less than 0.3 for all candidate parking spaces in the set during the calculation with A, it is considered that target A is a new parking space, and the coordinates of A, the occupancy status (x a1 , y a1 , x a2 , y a2 , 1), and the number of detection times are added to the set.

[0079] When the IoU between A and a certain candidate parking space B in the set is ≥ 0.3, that is, target A is parked at the position of the previous candidate B, it is considered that target A and candidate B are duplicate parking spaces, and the coordinates of A and B are weighted and averaged to obtain the updated upper-left coordinates:

[0080]

[0081] And the updated lower-right coordinates:

[0082]

[0083] Among them, n is the number of detection times of B, and the updated number of detection times is n + 1. Delete B from the set Φ, and add the updated coordinates, occupancy status, and number of detection times to the set Φ, that is, the position of the parking space is corrected using A.

[0084] (3) When the number of elements in the set Φ reaches the set number of parking spaces and does not increase within a certain period of time, the element coordinates in the set Φ can be considered as the coordinates of all parking spaces, and the parking space detection is completed.

[0085] In the above method, when the state is updated after the vehicle leaves the parking space: the states of these parking spaces are identified as "idle" and assigned 0.

[0086] In this way, the system continuously adjusts and updates the coordinates and states of the parking spaces by detecting and comparing the positions of the vehicles. Each time a new vehicle enters or leaves, the system dynamically updates the status (idle or occupied) and position of the parking space, making the position of the parking space more reasonable and accurate. This dynamic adjustment mechanism can ensure that the system reflects the actual status of each parking space in the parking lot in real time.

[0087]

[0088] For the details of the algorithm, see Algorithm 1. Based on the vehicle detection-based parking space detection algorithm, in the parking lot environment with a large traffic flow, this method demonstrates excellent performance. As long as each parking space is occupied once, all parking spaces can be detected, so that the positions of all parking spaces within the entire monitoring range can be automatically obtained in a short time. Its simple and efficient characteristics make it an ideal choice for automatically detecting parking spaces in high-traffic parking lots.

[0089] The second aspect

[0090] The present application provides a parking space monitoring system based on vehicle detection, including the following units: an image acquisition unit, a parking space set judgment unit, a comparison unit, and a parking space status monitoring unit;

[0091] The image acquisition unit is used to determine the parking position coordinates of the currently entering vehicle by acquiring images through a camera installed in the parking lot;

[0092] The parking space set judgment unit is used to judge whether the parking space set is empty; if it is empty, add the parking position coordinates of the currently entering vehicle, the occupied identification status of this parking space, and the number of times n that this parking space is detected to the parking space set as a pre-stored parking space. If it is not empty, execute the comparison unit; where the initial value of n is 1;

[0093] The comparison unit is used to compare the parking position coordinates of the currently entering vehicle with the coordinates of each pre-stored parking space in the parking space set;

[0094] If the position overlap between the rectangular frames represented by all coordinates is less than the low preset threshold, it is determined that the parking position coordinates of the currently entering vehicle are a new parking space, and the parking position coordinates of the currently entering vehicle, the occupied identification status of this parking space, and the number of times n that this parking space is detected are added to the parking space set as a pre-stored parking space; where the initial value of n is 1;

[0095] If the position overlap between the rectangular frames represented by two coordinates reaches the high preset threshold, it is determined that the parking position of the currently entering vehicle and this compared pre-stored parking space are duplicate parking spaces, correct the parking position coordinates of the currently entering vehicle, and add the corrected coordinates, the occupied identification status of this parking space, and the updated number of times the parking space is detected to the parking space set as a pre-stored parking space. After that, delete this compared pre-stored parking space;

[0096] When the number of pre-stored parking spaces in the parking space set reaches the preset number of parking spaces and there is no increase within the preset time period, the pre-stored parking space coordinates in the parking space set are the coordinates of all parking spaces;

[0097] The parking space status monitoring unit is used to judge the occupancy status of the parking space based on all the parking space coordinates obtained by the comparison unit.

[0098] Wherein, it further includes a status modification unit, and the status modification unit is used to modify the occupied identification status of a pre-stored parking space to an unoccupied identification status when it is detected that the vehicle on a certain pre-stored parking space leaves the parking lot.

[0099] Among them, the image acquisition unit is further configured to perform vehicle detection on the image using a trained vehicle detection model. The original image dataset used for training the vehicle detection model is generated by adding rain, snow, and fog weather effects through an artificially simulated weather effect algorithm to generate a simulated image dataset.

[0100] Among them, the system can be automatically called and executed through an automated script tool.

[0101] In addition, the present invention provides a storage medium, in which at least one instruction, at least one program, a code set or an instruction set is stored, and the at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by a processor to implement the above-mentioned parking space monitoring method based on vehicle detection.

[0102] In addition, the present invention provides a terminal, which includes a processor and a memory. At least one instruction, at least one program, a code set or an instruction set is stored in the memory, and the at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by the processor to implement the above-mentioned parking space monitoring method based on vehicle detection.

[0103] The optional implementation manners of the embodiments of the present invention have been described in detail above in conjunction with the accompanying drawings. However, the embodiments of the present invention are not limited to the specific details in the above implementation manners. Within the scope of the technical concept of the embodiments of the present invention, various simple modifications can be made to the technical solutions of the embodiments of the present invention, and these simple modifications all fall within the protection scope of the embodiments of the present invention.

Claims

1. A parking space monitoring method based on vehicle detection, characterized in that, It includes the following steps: Step S1: Determine the parking position coordinates of the currently entering vehicle by obtaining an image through a camera installed in the parking lot; Step S2: Judge whether the parking space set is empty; If it is empty, add the parking position coordinates of the currently entering vehicle, the occupied flag status of this parking space, and the detection times n of this parking space to the parking space set as a pre-stored parking space. If it is not empty, jump to Step S3; where the initial value of n is 1; Step S3: Compare the parking position coordinates of the currently entering vehicle with the coordinates of each pre-stored parking space in the parking space set; If the position overlap between the rectangular frames represented by all coordinates is less than the low preset threshold, it is determined that the parking position coordinates of the currently entering vehicle are a new parking space, and add the parking position coordinates of the currently entering vehicle, the occupied flag status of this parking space, and the detection times n of this parking space to the parking space set as a pre-stored parking space; where the initial value of n is 1; If the position overlap between the rectangular frames represented by two coordinates reaches the high preset threshold, it is determined that the parking position of the currently entering vehicle and the compared pre-stored parking space are duplicate parking spaces, correct the parking position coordinates of the currently entering vehicle, and add the corrected coordinates, the occupied flag status of this parking space, and the updated detection times of the parking space to the parking space set as a pre-stored parking space. After that, delete the compared pre-stored parking space; When the number of pre-stored parking spaces in the parking space set reaches the preset number of parking spaces and there is no increase within the preset time period, the pre-stored parking space coordinates in the parking space set are all the parking space coordinates; Step S4: Judge the occupancy status of the parking space based on all the parking space coordinates obtained in Step S3.

2. The method for monitoring a parking space based on vehicle detection according to claim 1, wherein It also includes the step of modifying the occupied flag status of a pre-stored parking space to an unoccupied flag status when it is detected that the vehicle on a certain pre-stored parking space leaves the parking lot.

3. The parking space monitoring method based on vehicle detection according to claim 1, wherein, In Step S1, it includes the step of using a trained vehicle detection model to detect vehicles in the image. Among them, the original image data set used for training the vehicle detection model generates a simulated image data set by adding rain, snow, and fog weather effects through an artificial simulation weather effect algorithm.

4. The method for monitoring a parking space based on vehicle detection according to claim 3, wherein, The method can be automatically called and executed through an automated script tool.

5. A parking space monitoring system based on vehicle detection, characterized in that, It includes the following units: an image acquisition unit, a parking space set judgment unit, a comparison unit, and a parking space status monitoring unit; The image acquisition unit is used to determine the parking position coordinates of the currently entering vehicle by obtaining an image through a camera installed in the parking lot; The parking space set judgment unit is used to judge whether the parking space set is empty; If it is empty, add the parking position coordinates of the currently entering vehicle, the occupied flag status of this parking space, and the detection times n of this parking space to the parking space set as a pre-stored parking space. If it is not empty, execute the comparison unit; where the initial value of n is 1; The comparison unit is used to compare the parking position coordinates of the currently entering vehicle with the coordinates of each pre-stored parking space in the parking space set; If the position overlap between the current entering vehicle's parking position coordinates and the rectangular frames represented by all coordinates is less than a low preset threshold, the current entering vehicle's parking position coordinates are recognized as a new parking space, and the current entering vehicle's parking position coordinates, the occupied identification status of this parking space, and the number of times n that this parking space has been detected are added to the parking space set as a pre-stored parking space; where the initial value of n is 1; If the position overlap between the rectangular frames represented by two coordinates reaches a high preset threshold, it is recognized that the current entering vehicle's parking position and this compared pre-stored parking space are duplicate parking spaces, the current entering vehicle's parking position coordinates are corrected, and the corrected coordinates, the occupied identification status of this parking space, and the updated number of times the parking space has been detected are added to the parking space set as a pre-stored parking space. After that, this compared pre-stored parking space is deleted; When the number of pre-stored parking spaces in the parking space set reaches the preset number of parking spaces and there is no increase within a preset time period, the pre-stored parking space coordinates in the parking space set are all the parking space coordinates; The parking space status monitoring unit is used to judge the occupied status of the parking space based on all the parking space coordinates obtained by the comparison unit.

6. The parking space monitoring system based on vehicle detection according to claim 5, wherein It further includes a status modification unit, which is used to modify the occupied identification status of a certain pre-stored parking space to an unoccupied identification status when it is detected that the vehicle on this pre-stored parking space has left the parking lot.

7. The parking space monitoring system based on vehicle detection according to claim 5, characterized in that, The image acquisition unit is further used to perform vehicle detection on the image using a trained vehicle detection model, where the original image data set used for training the vehicle detection model is generated by adding rain, snow, and fog weather effects through an artificially simulated weather effect algorithm to generate a simulated image data set.

8. The parking space monitoring system based on vehicle detection according to claim 7, characterized in that, The system can be automatically called and executed through an automated script tool.

9. A storage medium, characterized in that, At least one instruction, at least one program, a code set, or an instruction set is stored in the storage medium, and the at least one instruction, the at least one program, the code set, or the instruction set is loaded and executed by a processor to implement the vehicle detection-based parking space monitoring method according to any one of claims 1-4.

10. An electronic terminal, characterized in that, The electronic terminal includes a processor and a memory, and at least one instruction, at least one program, a code set, or an instruction set is stored in the memory. The at least one instruction, the at least one program, the code set, or the instruction set is loaded and executed by the processor to implement the vehicle detection-based parking space monitoring method according to any one of claims 1-4.

Images