WO2023179416A1

WO2023179416A1 - Method and apparatus for determining entry and exit of vehicle into and out of parking space, device, and storage medium

Info

Publication number: WO2023179416A1
Application number: PCT/CN2023/081494
Authority: WO
Inventors: 神克乐; 陈新; 周浩; 荆碧晨; 徐博文; 龙一民; 邵懿; 杜伟
Original assignee: 阿里云计算有限公司
Priority date: 2022-03-22
Filing date: 2023-03-15
Publication date: 2023-09-28
Also published as: CN114694095A

Abstract

The present application provides a method and an apparatus for determining entry and exit of a vehicle into and out of a parking space, a device, and a storage medium. The method comprises: acquiring a first image collected by a camera, and determining positions of a vehicle detection frame and a preset vehicle body center point of a target vehicle in the first image; if it is determined that the positions of the vehicle detection frame and the preset vehicle body center point meet a parking space occupation condition of a target parking space border, acquiring vehicle tracking information of the target vehicle in multiple second images collected within a preset time before acquiring the first image, and if it is determined, according to the vehicle tracking information, that the target vehicle is within an early warning border, determining that the target vehicle enters a parking space corresponding to the target parking space border to park. By means of defining an early warning region outside multiple parking spaces covered by a camera, so as to assist in determining entry and exit of a vehicle into and out of a parking space, and by means of dual determination of a parking space occupation condition and an early warning region passed through, the accuracy of a result of determining entry and exit into and out of a parking space is improved.

Description

Determine the methods, devices, equipment and storage media for vehicles entering and exiting parking spaces

This application requires the priority of the Chinese patent application submitted to the China Patent Office on March 22, 2022, with the application number 202210288663.2, and the application name is "Method, device, equipment and storage medium for determining vehicle entry and exit from parking spaces", and its entire content has been approved This reference is incorporated into this application.

Technical field

The present invention relates to the technical field of image processing, and in particular to a method, device, equipment and storage medium for determining the entry and exit of a vehicle into a parking space.

Background technique

As my country's urbanization process accelerates, the growth rate of vehicles is also increasing day by day. The problem caused by this is the difficulty of urban parking. For city managers, optimizing parking resources and improving the level of intelligent parking management have become very important goals in the development of new smart cities. Therefore, automated toll management of roadside parking spaces is particularly important.

Automated charging strategies for roadside parking spaces currently include the following methods: charging with handheld terminal devices, installing geomagnetism, and installing video piles. Among them, the charging method of handheld terminal equipment requires manual labor to carry the terminal equipment continuously for duty, and there are omissions and other phenomena. It also relies heavily on manual labor and low technology, resulting in high operating costs. This method is obviously not suitable for today's smart city technology. Target. The second is the geomagnetic solution, which uses sensors within the geomagnetic field to sense whether there are obstacles above the geomagnetic field. Since it is impossible to determine whether the obstacle is a parking vehicle or other non-motorized vehicle, this method is susceptible to interference and has a high false alarm rate.

The premise of the automated charging strategy for roadside parking spaces is to be able to automatically and accurately identify vehicles entering and leaving the parking spaces. Therefore, how to complete the automatic and accurate identification of vehicles entering and parking at the berth and leaving the berth is a problem that needs to be solved.

Contents of the invention

Embodiments of the present invention provide a method, device, equipment and storage medium for determining vehicle entry and exit from a berth, so as to improve the accuracy of the determination result of a vehicle's entry and exit from a berth.

In a first aspect, an embodiment of the present invention provides a method for determining whether a vehicle enters or exits a parking space. The method includes:

Obtaining a first image collected by a camera covering multiple berths;

Perform vehicle detection processing on the first image to obtain the vehicle detection frame corresponding to the target vehicle in the first image and the position of the preset center point of the vehicle body;

If it is determined that the positions of the vehicle detection frame and the preset center point of the vehicle body meet the parking space occupancy conditions of the target parking frame, then the vehicle of the target vehicle in the multiple frames of the second image collected within the preset time before the first image is acquired. Tracking information, the vehicle tracking information includes the corresponding positioning positions of the vehicle identification of the target vehicle in the multi-frame second images; wherein the target parking frame is the position of any of the parking spots in the camera. The corresponding image area boundary in the shooting screen;

If it is determined according to the vehicle tracking information that the vehicle identification appears in the preset warning frame, it is determined that the target vehicle enters the parking space corresponding to the target parking frame and stops; wherein the warning frame is the entrance to the parking space. The pre-warning area that multiple parking spaces need to pass through is the corresponding image area boundary in the shooting picture of the camera. The early warning area is a roadway area surrounding the multiple parking spaces.

In a second aspect, an embodiment of the present invention provides a device for determining the entry and exit of a vehicle into a parking space. The device includes:

An acquisition module, used to acquire the first image collected by a camera covering multiple berths;

A detection module configured to perform vehicle detection processing on the first image to obtain the vehicle detection frame corresponding to the target vehicle in the first image and the position of the preset center point of the vehicle body;

A processing module configured to obtain all the second images in multiple frames collected within a preset time before the first image if it is determined that the positions of the vehicle detection frame and the preset center point of the vehicle body meet the parking occupancy conditions of the target parking frame. The vehicle tracking information of the target vehicle, the vehicle tracking information includes the respective corresponding positioning positions of the vehicle identification of the target vehicle in the plurality of second images; and if the vehicle identification is determined according to the vehicle tracking information appears in the preset early warning frame, then it is determined that the target vehicle has entered the parking space corresponding to the target parking space frame and is parked; wherein, the early warning frame is the early warning area required to pass in and out of the multiple parking spaces. The corresponding image area boundary in the camera's shooting screen, the warning area is the roadway area surrounding the multiple parking spaces, and the target parking frame is the corresponding image area of any of the parking spaces in the camera's shooting screen. the image area boundary.

In a third aspect, embodiments of the present invention provide an electronic device, including: a memory, a processor, and a communication interface; wherein executable code is stored on the memory, and when the executable code is executed by the processor, The processor is enabled to at least implement the method for determining the vehicle entering and exiting the parking space as described in the first aspect.

In a fourth aspect, embodiments of the present invention provide a non-transitory machine-readable storage medium. The non-transitory machine-readable storage medium stores executable code. When the executable code is processed by a processor of an electronic device, When executed, the processor is enabled to at least implement the method for determining the entry and exit of a vehicle into a parking space as described in the first aspect.

In a fifth aspect, embodiments of the present invention provide a method for determining vehicle entry and exit from a roadside parking space, including:

Obtaining the first image collected by a roadside camera, the roadside camera includes a camera arranged on the same side of the roadside parking space, and the roadside camera covers multiple roadside parking spaces;

If it is determined that the positions of the vehicle detection frame and the preset center point of the vehicle body meet the parking space occupancy conditions of the target parking frame, then the vehicle of the target vehicle in the multiple frames of the second image collected within the preset time before the first image is acquired. Tracking information, the vehicle tracking information includes the corresponding position of the vehicle identification of the target vehicle in the multi-frame second image. position; wherein, the target parking frame is the boundary of the image area corresponding to any of the roadside parking spaces in the shooting screen of the roadside camera;

If it is determined according to the vehicle tracking information that the vehicle identification appears in the preset warning frame, it is determined that the target vehicle drives into the roadside parking space corresponding to the target parking frame and stops; wherein the warning frame is the entrance and exit The warning area that the multiple roadside parking spaces need to pass through is the corresponding image area boundary in the picture captured by the roadside camera. The warning area is a roadway area surrounding the multiple roadside parking spaces.

For the situation where a vehicle enters a parking space set up on the roadside to park, a camera can be set up on the roadside on the same side of the roadside parking space, called a roadside camera. One roadside camera can be configured to cover multiple fixed roadside parking spaces. The roadside camera cannot be rotated, so the roadside camera has a fixed shooting range. That is to say, for the multiple roadside parking spaces that can be covered by one roadside camera, these multiple roadside parking spaces can be covered by the roadside camera. The captured image always corresponds to a fixed image area, and the closed curve enclosed by the boundaries of the image area is called the berth border. For multiple roadside parking spaces covered by a roadside camera, an early warning area is defined on the same side of the roadside parking space. The early warning area is the roadway area surrounding the multiple roadside parking spaces. Vehicles entering and exiting these multiple roadside parking spaces can Roadside parking spaces must pass through this warning area. Based on this definition, it can be understood that the warning area also has a fixed image area in the picture captured by the roadside camera, and the closed curve formed by the boundary of the image area is called a warning frame.

In the embodiment of the present invention, for any camera (such as the above-mentioned roadside camera), the camera takes a video picture, and the video picture captured by the camera can be sampled to obtain a frame of image. For each frame of image obtained by sampling, , can perform vehicle detection and vehicle tracking processing. The purpose of vehicle detection is to detect the vehicle detection frame of each vehicle contained in the image and the position of a certain center point preset on the body. This position refers to the position of the center point in the image. corresponding pixel position in . The purpose of vehicle tracking is to identify the same vehicle in different frame images and determine the vehicle's vehicle identity. The vehicle identification can be a recognized license plate number or the same tracking number assigned to the same vehicle.

For any frame of image collected by the above camera (called the first image), vehicle detection processing is performed on the first image to obtain the vehicle detection corresponding to the target vehicle (referring to any vehicle included in it) in the first image. After the position of the center point of the frame and the body is preset, and on the premise that the multiple parking spaces covered by the camera are known to have corresponding parking frame frames in the image, first, determine whether the target vehicle meets the parking occupancy conditions, that is, combined with the detected The vehicle detection frame of the target vehicle and the position of the preset center point on the vehicle body determine whether the target vehicle currently occupies a parking space corresponding to a target parking frame. If it is determined that the positions of the above-mentioned vehicle detection frame and the preset center point of the vehicle body meet the parking occupancy conditions of the target parking frame, the time will be traced back to obtain the second image of multiple frames collected within the preset time (such as 10 seconds) before the first image. Vehicle tracking information of the target vehicle. The vehicle tracking information includes the corresponding positioning positions of the vehicle identification of the target vehicle in the multiple second images. That is to say, based on the previous vehicle tracking process, the target vehicle corresponding to the same vehicle identification can be obtained. The position where the frame appears in the image. Further, it is determined according to the obtained vehicle tracking information whether the target vehicle that entered the parking space corresponding to the target parking frame has appeared in the warning area before entering the parking space, that is, it is determined whether the vehicle identification of the target vehicle appears in the second image of the multiple frames. Within the warning frame, if the vehicle logo appears in one or several frames of the second image, it is finally determined that the target vehicle enters the parking space corresponding to the target parking space frame and stops.

In the above solution, an early warning area is defined outside several consecutive parking spaces covered by the same camera to assist in the judgment of vehicles entering and exiting the parking spaces. Only after it is first determined that the vehicle meets the conditions for occupying the parking space, it is determined that the vehicle meets the requirements of the early warning area. When the conditions are met, it is finally determined that the vehicle has entered the parking space and parked, thereby improving the accuracy of the entry determination result. Moreover, when determining parking space occupancy, combining the two factors of the vehicle detection frame and a specific center point on the vehicle body to determine whether the vehicle occupies a certain parking space can improve the accuracy of the parking space occupancy determination result.

Description of the drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings without exerting creative efforts.

Figure 1 is a schematic diagram of a berth parking scene provided by an embodiment of the present invention;

Figure 2 is a flow chart of a method for determining that a vehicle has entered a parking space according to an embodiment of the present invention;

Figure 3 is a schematic diagram of a vehicle detection frame and a vehicle chassis detection frame provided by an embodiment of the present invention;

Figure 4 is a schematic diagram of a vehicle detection frame and a vehicle chassis center point determination principle provided by an embodiment of the present invention;

Figure 5 is a flow chart of a method for determining that a vehicle moves out of a parking space according to an embodiment of the present invention;

Figure 6 is a schematic structural diagram of a device for determining vehicle entry and exit from a parking space provided by an embodiment of the present invention;

FIG. 7 is a schematic structural diagram of an electronic device provided by the embodiment shown in FIG. 6 .

Detailed ways

In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments These are some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without making creative efforts fall within the scope of protection of the present invention.

Some embodiments of the present invention will be described in detail below with reference to the accompanying drawings. The following embodiments and features in the embodiments may be combined with each other as long as there is no conflict between the embodiments. In addition, the sequence of steps in the following method embodiments is only an example and is not strictly limited.

The method for determining vehicle entry and exit parking spaces provided by the embodiment of the present invention is suitable for parking management application scenarios of roadside parking spaces and indoor and outdoor parking lots. Roadside parking spaces refer to the allowed spaces demarcated on the side of the roadway close to the sidewalk. Parking space.

In the traditional manual and geomagnetic berth charging management solution, in addition to the large labor cost, each berth needs to independently set up a geomagnetic device, which also has a large equipment overhead. Moreover, this solution also didn't stop Car pictures or video evidence links can easily lead to parking fee disputes. For this reason, for roadside parking spaces, the video pile solution can be used. The video pile solution mainly installs a camera on the road sidewalk at the same height as the vehicle, close to the roadside parking space. Although this solution can retain the parking evidence chain, however, The maintenance cost is high and it is easy to be damaged manually. In addition, the roads in many northern cities have a lot of dust, and the dust can easily block the camera within 3 meters of the road. Therefore, manual cleaning is also required on a regular basis.

In view of the disadvantages of the above solutions, in the case of roadside parking spaces, the embodiment of the present invention proposes a high-level video solution. This solution requires the installation of high poles with higher heights (generally more than 6 meters) on the sidewalks of the roads, and then Cameras are installed on high poles. Each camera is responsible for photographing several (usually 3-4) parking spaces, and uses artificial intelligence algorithms to identify vehicles entering and exiting the parking spaces and generate corresponding entry and exit events, such as recording the parking start of vehicles. and end time, as well as the corresponding vehicle identification (such as license plate number), and save the picture/video evidence chain of the vehicle entering and exiting, so as to realize automatic vehicle toll collection management for roadside parking spaces. This solution can effectively solve the problems of previous solutions that are susceptible to interference, have no evidence chain, and have high manual operation and maintenance costs. Moreover, high poles can be installed by borrowing poles (such as borrowing street light poles), making rational use of existing municipal facilities. Reduce impact on urban landscape. Similarly, for indoor and outdoor parking lots, cameras can also be set up at higher locations.

For the convenience of description, the embodiment of the present invention takes the roadside parking scene as an example for description. For a roadside parking space designated on a certain side of the road, the camera used to take pictures of the roadside parking space (called a roadside camera) can be set on the opposite side of the roadside parking space or on the roadside. Same side of the berth. For example, if there is a row of roadside parking spaces defined on the left edge of a certain road, the roadside camera can be located on the sidewalk on the left side of the road, that is, on the same side as the roadside parking spaces, or it can be located on the sidewalk on the right side of the road, that is, opposite to the roadside parking spaces. side settings. Of course, in actual applications, roadside cameras may exist on both the same side and the opposite side.

It should be noted that a roadside camera has a limited shooting range. Therefore, a roadside camera is generally configured to cover several roadside parking spaces. Moreover, the roadside camera can be configured not to rotate, that is, to have a fixed shooting angle. In this way, multiple roadside parking spaces covered by the same roadside camera have a fixed image area in all the images captured by the roadside camera. .

When the roadside camera is set up on the same side as the roadside parking space, as shown in Figure 1, the blocking of vehicles in the front and rear parking spaces is often serious, and the blocking rate is generally greater than 60%, and sometimes even close to 100%. Therefore , in the case of roadside cameras installed on the same side as the roadside parking space, it is very challenging to accurately identify vehicles entering and exiting the parking space. The vehicles entering and exiting the berth mentioned here refer to vehicles that actually drive into the berth for parking and then drive out of the berth after parking for a period of time.

In the embodiment of the present invention, in order to accurately identify vehicles entering and exiting the parking space, an early warning area is defined. As shown in Figure 1, on the roadway in the same direction outside the area occupied by multiple roadside parking spaces covered by the same roadside camera, a roadway area surrounding these multiple roadside parking spaces is set, which is called an early warning. area. This warning area represents the area that all vehicles must pass through from appearing in the roadside camera's picture to entering one of the roadside parking spaces or leaving the roadside parking space. As shown in Figure 1, the early warning area "surrounds" multiple roadside berths. It does not mean that multiple roadside berths are located inside the early warning area (belonging to an inclusion relationship), but from a positional relationship, the early warning area and multiple roadside berths are The two roadside berths are separate, that is, there is no inclusion or intersection relationship.

It is understandable that there is no need to actually draw this warning area on a real roadway. This warning area is only an area defined to assist in accurately identifying vehicles entering and exiting the parking space. Since the multiple roadside areas it surrounds have fixed image areas in the pictures captured by the corresponding roadside cameras, and the positional relationship between the warning area and the multiple roadside areas is also fixed, therefore, the warning The area also has a fixed image area in the picture captured by the roadside camera.

In the embodiment of the present invention, the boundary of the image area corresponding to the roadside parking space in the shooting screen of the corresponding roadside camera is called the parking frame, and the boundary of the image area corresponding to the warning area in the shooting screen of the roadside camera is called For the warning border.

The parking space frame may be based on the actual roadside parking boundary line drawn on the roadway in the picture taken by the roadside camera in advance when no vehicle enters the multiple roadside parking spaces covered by the roadside camera. The image area is marked.

Among them, the early warning frame can be based on the spatial positional relationship between the predefined early warning area and multiple roadside parking spaces (that is, the position and distance relationship between the two in real road scenes), as well as the images captured by roadside cameras of multiple roadside parking spaces. The corresponding image area in , the image area of the warning area in the picture captured by the roadside camera is mapped.

The method for determining whether a vehicle enters or exits a roadside parking space provided by the embodiment of the present invention can be executed by an electronic device. The electronic device can be a server or terminal device that is communicatively connected to the roadside camera. The server can be a physical server or a virtual server in the cloud. (virtual machine). Of course, during the actual execution of the method provided by the embodiment of the present invention, the roadside camera can also be completed by cooperating with the electronic device. For example, the vehicle detection, vehicle tracking and other processing of the image can also be completed locally on the roadside camera. The electronic device receives data transmitted by the roadside camera related to generating the parking record in order to generate and store the parking record.

The implementation process of the solution for determining vehicle entry and exit from a roadside parking space provided by the embodiment of the present invention is described below.

Figure 2 is a flow chart of a method for determining that a vehicle has entered a parking space according to an embodiment of the present invention. As shown in Figure 2, the method includes the following steps:

201. Obtain the first image collected by the roadside camera. The roadside camera includes a camera installed on the same side of the roadside parking space. The roadside camera covers multiple roadside parking spaces.

202. Perform vehicle detection processing on the first image to obtain the vehicle detection frame corresponding to the target vehicle in the first image and the position of the preset center point of the vehicle body.

203. If it is determined that the positions of the vehicle detection frame and the preset center point of the vehicle body meet the parking occupancy conditions of the target parking frame, obtain the vehicle tracking information of the target vehicle in the second image collected within the preset time before the first image. , the vehicle tracking information includes the corresponding positioning positions of the vehicle identification of the target vehicle in the plurality of second images.

Wherein, the target parking frame is the boundary of the image area corresponding to one of the plurality of roadside parking spaces in the picture captured by the roadside camera.

204. If it is determined according to the vehicle tracking information that the vehicle identification appears within the preset warning frame, then determine that the target vehicle drives into the roadside parking space corresponding to the target parking space frame and stops.

Wherein, as mentioned above, the early warning frame is the boundary of the image area corresponding to the early warning area required to enter and exit the multiple roadside parking spaces in the shooting screen of the roadside camera, and the early warning area is surrounding the multiple roadside parking spaces. The roadway area of the parking space.

For the roadside camera X covering the above-mentioned multiple roadside parking spaces, the roadside camera X can continuously capture video images. Faced with the need to accurately identify vehicles entering and exiting roadside parking spaces, the video footage captured by the roadside camera Track processing. Among them, the sampling frequency can be preset, such as 5 frames/second.

Among them, the purpose of vehicle detection is to detect the vehicle detection frame of each vehicle contained in the image and the position of a certain center point preset on the vehicle body. This position refers to the corresponding pixel position in the image. The purpose of vehicle tracking is to identify the same vehicle in different frame images and determine the vehicle's vehicle identity. The above-mentioned preset center point may be the center point of the vehicle chassis.

For vehicle detection processing, the above-mentioned first image can be any frame image collected by the roadside camera Vehicle detection processing can be implemented as:

Use the pre-trained deep neural network model to perform vehicle detection processing on the first image to obtain the vehicle detection frame and vehicle chassis detection frame corresponding to the target vehicle in the first image, and determine the position of the center point of the vehicle chassis detection frame as the vehicle The position of the center point of the chassis. The target vehicle refers to any vehicle detected from the first image.

In the embodiment of the present invention, the vehicle detection frame refers to the detection frame including the complete vehicle, and the vehicle chassis detection frame refers to the detection frame including the vehicle chassis area, as shown in Figure 3.

The vehicle detection frame and the vehicle chassis detection frame are both a rectangular frame. When the vehicle chassis detection frame is obtained, the pixel coordinates of its four vertices in the image can be obtained, and then the pixels of the center point can be obtained based on the pixel coordinates of the four vertices. The coordinate is the position of the center point of the vehicle chassis.

In fact, vehicle detection is also a target detection task for specific targets such as vehicles in images. Therefore, the above-mentioned deep neural network model can be trained using the training ideas of the target detection task. Simply put, a large number of training samples are collected in advance, which can include training sample images in which the vehicle presents various postures and occlusion situations in the image. By annotating the supervision information in the training sample images, the depth is carried out based on the annotated supervision information. Training of neural network models. Among them, the annotated supervision information is the vehicle detection frame and vehicle chassis detection frame corresponding to the vehicle in the corresponding training sample image.

In fact, when the roadside camera is installed on the same side as the roadside parking space, from the perspective of the roadside camera, there may be serious occlusion between the vehicles parked on the roadside parking space, as shown in Figure 1 As shown, the vehicle with license plate A blocks most of the vehicle with license plate B. At this time, it may be unreliable to rely solely on the detection results of the above-mentioned deep neural network model to determine the location of the chassis center point, because the deep neural network model at this time The vehicle chassis detection frame may not be output.

For this situation, optionally performing vehicle detection processing on the first image can also be implemented as:

Use the pre-trained deep neural network model to perform vehicle detection processing on the first image to obtain the vehicle detection frame corresponding to the target vehicle in the first image;

Determine the sub-detection frame with a preset height ratio at the bottom of the vehicle detection frame, where the preset height ratio is set based on the height of the vehicle chassis from the ground;

The position of the top center point of the sub-detection frame is determined as the position of the center point of the vehicle chassis.

That is to say, at this time, the deep neural network model can only be trained to identify the vehicle detection frame, and the corresponding position coordinates of the vehicle chassis center point in the first image are indirectly determined based on the obtained vehicle detection frame.

For ease of understanding, an exemplary description is provided with reference to Figure 4. In Figure 4, it is assumed that the deep neural network model detects a vehicle detection frame Q illustrated in the figure from the first image, and it is assumed that the above-mentioned preset height ratio is 1/5 , that is, the height H of the vehicle detection frame is divided into 5 parts, and the bottom 1/5H forms a sub-detection frame q. The center point of the top width of the sub-detection frame q (the circle point in the picture) is the chassis center point. This height of 1/5H is obtained by counting the ratio of the height of the vehicle chassis from the ground to the height of the top of the vehicle from the ground.

In the embodiment of the present invention, in addition to performing vehicle detection processing on each frame of image sampled to obtain the vehicle detection frame corresponding to each vehicle contained in each frame of image and the position of the center point set on the vehicle body, the Continuous different frame images are processed for vehicle tracking. Because the vehicle detection process can only determine whether a frame of image contains a vehicle and the corresponding position of the contained vehicle in the image, it cannot know the identification information of the vehicle and the movement trajectory information of the vehicle, so it cannot know the different frame images. Identity between vehicles. Through vehicle tracking processing, the identification information of the vehicle can be obtained and the corresponding position of the same vehicle in different frame images can be determined through tracking.

Vehicle detection processing and vehicle tracking processing can cooperate with each other. For example, when a vehicle is found in a frame of image through vehicle detection processing and the corresponding vehicle detection frame is obtained, vehicle tracking processing can identify the visual characteristics of the vehicle (such as color, model, etc.) , size, outline, etc.), combined with characteristic information such as moving speed, you can know whether this vehicle is the same vehicle as a vehicle that appeared in some previous images. Vehicle tracking processing can be implemented with reference to existing related technologies, and will not be described in detail in this embodiment.

In practical applications, vehicle detection processing and vehicle tracking processing can be executed by different processes. When a vehicle just appears in the picture of the roadside camera X, the vehicle detection processing process and the vehicle tracking processing process detect the vehicle for the first time. For a vehicle, the vehicle tracking process will assign a tracking number as a kind of identification information for the vehicle. At the same time, the vehicle tracking processing process can also perform license plate recognition processing to identify the license plate number of the vehicle. If the license plate number can be recognized (actually, it may be due to factors such as long distance, fast moving speed, occlusion between vehicles, etc.). If the license plate number cannot be recognized), the license plate number and the tracking number can be used as the identification information of the vehicle.

Optionally, during specific implementation, if the vehicle detection processing process detects a vehicle in a certain frame of image, the vehicle tracking processing process will extract the characteristics of the vehicle such as visual features, moving speed, etc. Once the vehicle identification is determined, the vehicle identification can be associated with the corresponding vehicle detection frame. In this way, through the joint processing of vehicle detection and vehicle tracking, it is possible to know the corresponding vehicle detection frame and vehicle identification of a vehicle in different frame images. Therefore, based on the same vehicle identification, the corresponding vehicle in multiple consecutive frames of images can be determined. The position of the detection frame can form the movement trajectory of the car.

The above introduces the process of vehicle detection and vehicle tracking processing for each frame of image collected by the roadside camera X. Continuing from the above, the vehicle detection frame corresponding to the target vehicle detected in the first image and the preset center point of the body (such as For example, in order to determine whether the target vehicle has parked at a certain roadside parking space covered by the roadside camera The position is compared with the corresponding parking frame of the multiple roadside parking spaces covered by the roadside camera

The parking space occupancy conditions include: the coincidence degree between the vehicle detection frame and the target parking frame is greater than the set threshold, and the position of the center point of the vehicle chassis is within the target parking frame. Wherein, the target parking frame is any one of the parking frame corresponding to the above-mentioned multiple roadside parking spaces. If the above conditions are met, it is determined that the target vehicle occupies the roadside parking space corresponding to the target parking frame. The set threshold corresponding to the above-mentioned degree of coincidence is, for example, a preset value such as 30%.

In practical applications, the coincidence degree of the vehicle detection frame and the target parking frame can be judged first. If the coincidence degree is greater than the set threshold, then it is judged whether the center point of the vehicle chassis is within the target parking frame. Only these two conditions are met. , the subsequent judgment process is executed. If the degree of coincidence is less than the above set threshold, no subsequent judgment will be made to determine that the target vehicle does not occupy the roadside parking space corresponding to the target parking frame.

In the embodiment of the present invention, through the dual judgment of the above-mentioned coincidence degree and the center point of the vehicle chassis, it can be more accurately determined whether the target vehicle occupies a roadside parking space, that is, a more accurate judgment of parking entry can be achieved. Because in fact, there may be situations where vehicles just pass by the roadside parking space instead of parking, and it is generally difficult for passing vehicles to meet the above two judgment conditions at the same time. For example, due to their relatively large bodies, large vehicles (such as buses and coaches) may meet the above-mentioned judgment conditions for coincidence when passing by roadside parking spaces, but they may not be able to meet the judgment conditions for the center point of the chassis.

When the vehicle detection frame of the target vehicle and the position of the center point of the vehicle chassis meet the parking occupancy conditions of the target parking frame, in order to more accurately determine that the target vehicle is a vehicle that drives into the corresponding roadside parking space, it is necessary to further determine whether the target vehicle is parked in the corresponding roadside parking space. Whether the target vehicle appeared in the warning area before entering the roadside parking space. If it has appeared in the warning area, it is finally determined that the target vehicle is a vehicle that drives into the roadside parking space corresponding to the target parking space frame, that is, the target vehicle has the behavior of driving into the roadside parking space and parking. If it has not appeared in the warning area, it is determined that the target vehicle does not drive into the roadside parking space and park.

In order to realize the above-mentioned judgment that the vehicle has appeared in the warning area, it is necessary to obtain vehicle tracking information of the target vehicle in multiple frames of second images collected within a preset time before the first image. The preset time is, for example, 10 seconds, 15 seconds and other preset values. The multi-frame second image refers to the multi-frame images obtained by sampling sequentially within this preset time. In order to distinguish it from the first image, it is called the multi-frame second image. image. Based on the above introduction to vehicle detection and vehicle tracking processing, it can be understood that the vehicle tracking information includes the corresponding positioning position of the vehicle identification of the target vehicle in the multiple second images. The positioning position may be in the corresponding second image. The position of the vehicle detection frame that hits the target vehicle. Of course, the position of a certain feature point (such as the center point, corner point) on the vehicle detection frame can also be used as the positioning position, and the positioning position is based on the vehicle tracking processing and The vehicle detection process has been obtained. If it is determined based on the vehicle tracking information that the vehicle identification of the target vehicle appears in the warning frame, it is determined that the target vehicle drives into the roadside parking space corresponding to the target parking frame and stops.

Among them, as mentioned above, the vehicle identification may include the license plate number and tracking number. During the judgment process, you can first determine whether the license plate number appears in the warning frame. If the license plate number is not successfully recognized due to long distance, fast vehicle movement, etc., you can then determine whether the tracking number appears in the warning frame.

Using the double confirmation method of license plate number and tracking number to determine whether the target vehicle appears in the warning area can improve the anti-interference of the determination result and overcome the interference of complex environments (such as occlusion, high-speed movement, and long distances).

To sum up, in the above scheme, an early warning area is defined outside several consecutive roadside parking spaces on the same side covered by the same roadside camera to assist in the judgment of vehicles entering and exiting the parking spaces. Only when it is first determined that the vehicle meets the requirements for occupying the roadside parking space After the conditions are met, it is determined that the vehicle meets the conditions for passing through the warning area before it is finally determined that the vehicle will enter the roadside parking space and park, thereby improving the accuracy of the entry determination result. Moreover, when determining parking space occupancy, combining the two factors of the vehicle detection frame and a specific center point on the vehicle body to determine whether the vehicle occupies a certain parking space can improve the accuracy of the parking space occupancy determination result.

In an optional embodiment, in order to further improve the accuracy of the judgment result of the vehicle entering the parking space, following the above example of the first image and the multi-frame second image, in combination with the target vehicle detected in the first image The vehicle detection frame and the center point of the vehicle chassis, as well as the positioning position of the target vehicle's vehicle identification in the corresponding second image tracked and determined from the multi-frame second image, determine that the target vehicle drove from the warning area into the road corresponding to the target parking frame. After parking in the roadside berth, it can be further determined whether the target vehicle will park stably in the roadside berth in the next period of time. If so, it is finally determined that the target vehicle has entered the roadside berth and parked. Otherwise, it is determined that the target vehicle is just passing by. The roadside berth.

Based on this, determining the roadside parking space corresponding to the target vehicle driving into the target parking frame can be implemented as:

Obtain vehicle tracking information of the target vehicle in multiple frames of third images collected within a preset time after the first image;

Carry out vehicle detection processing on multiple frames of third images respectively to obtain the positions of the target vehicle's corresponding vehicle detection frame and body preset center point in the multiple frames of third images by combining the vehicle tracking information of the target vehicle in the multiple frames of third images. ;

If the positions of the vehicle detection frame and the preset center point of the vehicle body respectively corresponding to the target vehicle in the multi-frame third image meet the parking occupancy conditions of the target parking frame, it is determined that the target vehicle drives into the roadside parking lot corresponding to the target parking frame and stops.

The preset time is, for example, 10 seconds, 15 seconds and other preset values. The multi-frame third image is similar to the multi-frame second image. It does not mean that there are multiple images of the same frame, but different frame images sampled respectively within the corresponding time period.

In summary, the vehicle tracking information here is mainly used to identify the same target vehicle. Carry out vehicle detection processing on multiple frames of the third image respectively to obtain the vehicle detection frame contained in each frame of image and the position of the preset center point of the vehicle body (such as the center point of the vehicle chassis), which can be combined with the vehicle tracking processing results to determine the location of the vehicle detection frame contained in each frame of image. The vehicle identification corresponding to the vehicle detection frame can be used to determine the vehicle detection frame of the target vehicle. If it is determined based on the corresponding vehicle detection frame and the position of the vehicle chassis center point of the target vehicle in the multi-frame third image that the target vehicle satisfies the conditions for occupying the roadside parking space corresponding to the target parking frame within the above preset time, then it is considered that The target vehicle is parked stably in the roadside parking space.

When it is finally determined that the target vehicle is parked at the roadside parking space corresponding to the target parking space frame, an entry record corresponding to the target vehicle may be generated. The entry record includes the identification of the roadside parking space corresponding to the target parking space frame, the target vehicle The vehicle identification, entry time and entry video, wherein the entry video at least includes video clips obtained by sampling multiple frames of the second image and the first image. Of course, it may also include video clips obtained by sampling multiple frames of the third image. Further, the entry record may also include the entry trajectory of the target vehicle: the entry trajectory is generated based on the positioning position corresponding to the vehicle identification of the target vehicle in the images obtained by sampling the entry video.

In an optional embodiment, in order to save computing power, the following berth occupancy condition determination strategy is provided:

For any sampled image i (such as the first image above), if the vehicle detection frame and the center point of the vehicle chassis of a vehicle (called the target vehicle) are detected, the vehicle detection frame is compared with the target When it is found that the vehicle detection frame is relatively far away from the target parking frame when the parking frame overlaps, the next image j for judging the parking occupancy conditions for the target vehicle can be determined based on the sampling frequency. The image j and the image i generally include samples. at least one frame of image. Assume that it is determined that the above judgment on the target vehicle needs to be made after m frames of images are separated, and m is greater than or equal to 1. Then based on vehicle tracking and vehicle detection processing, the vehicle detection frame and vehicle detection frame of the target vehicle are obtained in the image j after m frames are separated. After determining the center point of the chassis, compare the overlap between the vehicle detection frame and the target parking frame and determine whether the center point of the vehicle chassis is located within the target parking frame.

Since vehicle detection and vehicle tracking are both implemented based on deterministic deep neural network models and vehicle tracking algorithms, these two processes have stable computing power overhead. Judging parking space occupancy conditions also requires computing power overhead. Through the above strategy, we can Reduce the number of executions of this judgment, thereby reducing computing power overhead.

Figure 5 is a flow chart of a method for determining a vehicle to drive out of a parking space provided by an embodiment of the present invention. As shown in Figure 5, the method may include the following steps:

501. Perform vehicle detection processing on multiple fourth frames of images collected after the first image.

502. If it is determined based on the vehicle detection results of the fourth multi-frame image that the target vehicle does not meet the parking occupancy conditions of the target parking frame, obtain the vehicle tracking information of the target vehicle in the multi-frame fourth image.

503. If it is determined that the vehicle identification of the target vehicle appears in the warning frame based on the vehicle tracking information of the target vehicle in the fourth multi-frame image, then it is determined that the target vehicle has left the roadside parking space corresponding to the target parking frame.

The example taken from the previous embodiment is as follows: It is assumed that the target vehicle has been determined to drive into the roadside parking space corresponding to the target parking space frame and park based on the above-mentioned first image and multiple frames of second images collected before the first image.

It can be understood that when it is determined based on the multi-frame third image mentioned in the previous embodiment that the target vehicle drives into the roadside parking space corresponding to the target parking space frame and parks for a period of time, the first image in step 501 is replaced by the multi-frame third image. The last frame of the third image in the frame third image.

In summary, the "first image" in this step refers to the frame of image collected when the target vehicle is finally determined to drive into the roadside parking space corresponding to the target parking space frame for stable parking.

For the convenience of description, the roadside berth corresponding to the target berth frame is called the target roadside berth. Afterwards, in order to determine when the target vehicle drove out of the target roadside parking space, it is necessary to continue to perform vehicle detection and vehicle tracking processing on the images collected by the roadside camera X since the first image.

Assume that in the subsequent frame of image collected at T1, based on the vehicle detection results, it is found that the vehicle detection frame and the center point of the vehicle chassis that originally met the above parking occupancy conditions with the target parking frame no longer meet the parking occupancy conditions, that is, the vehicle detection frame If the coincidence degree with the target parking frame is lower than the set threshold, and/or the center point of the vehicle chassis is no longer within the target parking frame, it can be initially believed that the target vehicle originally parked at the target roadside parking space may have driven out of the target roadside parking space. . In order to ensure the reliability of the determination results, continue to observe whether the multi-frame images collected within a period of time after T1 (assuming the time reaches T2) are all in the vehicle detection frame and vehicle chassis of the target parking space where the target vehicle cannot be detected. If so, it is finally considered that the target vehicle does not meet the occupancy conditions of the target roadside parking space. In the above example, the multi-frame images collected from T1 to T2 are the above-mentioned multi-frame fourth images.

It can be understood that, as mentioned above, while vehicle detection processing is performed on these frame images, vehicle tracking processing is also performed, so that the vehicle detection frame corresponding to the target vehicle corresponding to the same vehicle identification in each frame image can be obtained and the vehicle chassis center point.

After determining that the target vehicle no longer satisfies the occupancy conditions of the target roadside parking space based on the vehicle detection results and vehicle tracking results of the fourth multi-frame image, the target vehicle is further determined based on the vehicle tracking information of the target vehicle obtained from the multi-frame fourth image. Whether it appears in the warning area during the time period from T1 to T2 is to determine whether the vehicle logo of the target vehicle has appeared within the warning frame in some of the images. The vehicle tracking information includes corresponding positioning positions of the vehicle identification of the target vehicle in the plurality of fourth images. If it has happened, it is determined that the target vehicle drove out of the target roadside parking lot. At this time, an exit record corresponding to the target vehicle can be generated. The exit record includes the identification of the target roadside parking space, the vehicle identification of the target vehicle, the exit time and A drive-out video, wherein the drive-out video at least includes video segments sampled to obtain the plurality of frames of fourth images. The corresponding relationship between the identification of the target roadside parking and its corresponding target parking frame is stored in advance and can be determined through query.

Since the roadside camera is set up on the same side as the roadside parking space, after the target vehicle parks at the target roadside parking space, due to the obstruction of other vehicles in front and behind it, it creates more difficulties to determine whether the target vehicle has left the target roadside parking space where it is parked. , therefore, in this embodiment, the dual confirmation method of the vehicle detection frame and the center point of the vehicle chassis is combined to determine whether the target vehicle no longer occupies the target roadside parking space, and at the same time, it is combined with whether the target vehicle passes the warning after not occupying the target roadside parking space. Regional judgment and multiple judgments can accurately determine the behavior of the target vehicle driving out of the target roadside parking space.

Based on the entry and exit records corresponding to the target vehicle, not only can the target vehicle be charged, but an accurate evidence chain can also be formed.

In each of the above embodiments, the scene of a vehicle entering and exiting a roadside parking lot is used as an example to illustrate the process of determining whether a vehicle enters or leaves the parking lot. As mentioned above, this solution can also be applied to indoor and outdoor parking lots. In other words, it is applicable to any scenario where there is a need for vehicles to enter and exit the parking space. In summary, in these indoor and outdoor parking scenes, one camera can be configured to capture multiple parking spaces (i.e., berths), and the warning area mentioned in the aforementioned embodiments can be set for each camera. Refer to the aforementioned implementation. In this example, the processing of video images collected by each camera can determine the entry and exit of vehicles in each parking space.

Based on this, in summary, the present invention provides a solution that is suitable for determining vehicles entering and exiting the parking space in any parking space scenario, as follows:

Obtaining a first image collected by a camera covering multiple berths;

If it is determined that the positions of the vehicle detection frame and the preset center point of the vehicle body meet the parking space occupancy conditions of the target parking frame, then the vehicle of the target vehicle in the multiple frames of the second image collected within the preset time before the first image is acquired. Tracking information, the vehicle tracking information includes the corresponding position of the vehicle identification of the target vehicle in the multi-frame second image. position; wherein, the target berth border is the boundary of the image area corresponding to any of the berths in the shooting screen of the camera;

For the specific implementation process of this solution, reference can be made to the relevant descriptions in other aforementioned embodiments and will not be described again here.

The device for determining the entry and exit of a vehicle into a parking space according to one or more embodiments of the present invention will be described in detail below. Those skilled in the art can understand that these devices can be constructed using commercially available hardware components and configured through the steps taught in this solution.

Figure 6 is a schematic structural diagram of a device for determining vehicle entry and exit from a parking space provided by an embodiment of the present invention. As shown in Figure 6, the device includes: an acquisition module 11, a detection module 12, and a processing module 13.

The acquisition module 11 is used to acquire the first image collected by a camera covering multiple berths.

The detection module 12 is configured to perform vehicle detection processing on the first image to obtain the vehicle detection frame corresponding to the target vehicle in the first image and the position of the preset center point of the vehicle body.

The processing module 13 is configured to, if it is determined that the positions of the vehicle detection frame and the preset center point of the vehicle body meet the parking occupancy conditions of the target parking frame, obtain the second image of the plurality of frames collected within the preset time before the first image. The vehicle tracking information of the target vehicle, the vehicle tracking information includes the corresponding positioning positions of the vehicle identification of the target vehicle in the multiple frames of second images; and if the vehicle is determined based on the vehicle tracking information If the logo appears in the preset warning frame, it is determined that the target vehicle has entered the parking space corresponding to the target parking space frame and is parked; wherein, the warning frame is where the warning area required to enter and exit the multiple parking spaces is located. The corresponding image area boundary in the shooting picture of the camera, the warning area is the roadway area surrounding the plurality of parking spaces, and the target parking frame is the location of any of the parking spaces in the shooting picture of the camera. The corresponding image area boundary.

Optionally, the parking space occupancy condition includes: the coincidence degree of the vehicle detection frame and the target parking space frame is greater than a set threshold, and the position of the preset center point of the vehicle body is located within the target parking space frame.

Optionally, the preset center point of the vehicle body includes the center point of the vehicle chassis.

Optionally, the detection module 12 is specifically configured to use a pre-trained deep neural network model to perform vehicle detection processing on the first image to obtain a vehicle detection frame corresponding to the target vehicle in the first image and The vehicle chassis detection frame refers to a detection frame containing a complete vehicle; the position of the center point of the vehicle chassis detection frame is determined as the position of the center point of the vehicle chassis.

Optionally, the detection module 12 is specifically configured to: use a pre-trained deep neural network model to perform vehicle detection processing on the first image to obtain a vehicle detection frame corresponding to the target vehicle in the first image; Determine the sub-detection frame with a preset height ratio at the bottom of the vehicle detection frame, and the preset height ratio is set according to the height of the vehicle chassis from the ground; determine the position of the top center point of the sub-detection frame as the The position of the center point of the vehicle chassis.

Optionally, the vehicle identification includes a license plate number and a tracking number. The tracking number is assigned during vehicle tracking processing of the image, and the same vehicle has the same tracking number.

Optionally, the processing module 13 is specifically configured to: obtain the vehicle tracking information of the target vehicle in multiple frames of third images collected within a preset time after the first image; Carry out vehicle detection processing to obtain vehicle detection frames and vehicle body preset center points respectively corresponding to the target vehicle in the multi-frame third image in combination with the vehicle tracking information of the target vehicle in the multi-frame third image. position; if the positions of the target vehicle's corresponding vehicle detection frame and body preset center point in the multi-frame third image meet the parking space occupancy conditions, then it is determined that the target vehicle has entered the target parking frame. Stop at the corresponding berth.

Optionally, the device further includes: a recording module, configured to generate an entry record corresponding to the target vehicle, the entry record including the identification of the parking space corresponding to the target parking space frame, the vehicle identification , the driving-in time and the driving-in video, wherein the driving-in video at least includes video segments sampling the plurality of frames of the second image and the first image.

Optionally, the detection module 12 is also configured to perform vehicle detection processing on multiple fourth frames of images collected after the first image. The processing module 13 is also configured to: if it is determined that the target vehicle does not meet the parking occupancy conditions of the target parking frame according to the vehicle detection results of the multi-frame fourth image, obtain all the information in the multi-frame fourth image. vehicle tracking information of the target vehicle; if it is determined that the vehicle identification appears in the warning frame according to the vehicle tracking information of the target vehicle in the fourth multi-frame image, then it is determined that the target vehicle has driven away from the warning frame. The berth corresponding to the target berth border.

Optionally, the recording module is also used to generate a drive-out record corresponding to the target vehicle. The drive-out record includes the identification of the parking space corresponding to the target parking space frame, the vehicle identification, the drive-out record time and a drive-out video, wherein the drive-out video at least includes a video segment sampling the plurality of frames of the fourth image.

The device shown in Figure 6 can perform the steps provided in the foregoing embodiments. For detailed execution processes and technical effects, please refer to the descriptions in the foregoing embodiments and will not be described again here.

In a possible design, the structure of the device for determining the vehicle entering and exiting the parking space shown in FIG. 6 can be implemented as an electronic device. As shown in FIG. 7 , the electronic device may include: a processor 21 , a memory 22 , and a communication interface 23 . The memory 22 stores executable code. When the executable code is executed by the processor 21, the processor 21 can at least implement the method for determining the vehicle entering and exiting the parking space as provided in the previous embodiment.

In addition, embodiments of the present invention provide a non-transitory machine-readable storage medium. The non-transitory machine-readable storage medium stores executable code. When the executable code is executed by a processor of an electronic device, , so that the processor can at least implement the method for determining the vehicle entering and exiting the parking space as provided in the previous embodiment.

The device embodiments described above are only illustrative, and the network elements described as separate components may or may not be physically separated. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Persons of ordinary skill in the art can understand and implement the method without any creative effort.

From the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by adding the necessary general hardware platform, or of course, can also be implemented by combining hardware and software. Based on this understanding, the above technical solution can be embodied in the form of a computer product in nature or in other words, the part that contributes to the existing technology. The present invention can use one or more computer-usable storage devices containing computer-usable program codes. The form of a computer program product implemented on media (including but not limited to disk storage, CD-ROM, optical storage, etc.).

Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that it can still be used Modifications are made to the technical solutions described in the foregoing embodiments, or equivalent substitutions are made to some of the technical features; however, these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

A method for determining the entry and exit of vehicles into a parking space, which is characterized by including:

Obtaining a first image collected by a camera covering multiple berths;

Perform vehicle detection processing on the first image to obtain the vehicle detection frame corresponding to the target vehicle in the first image and the position of the preset center point of the vehicle body;

If it is determined that the positions of the vehicle detection frame and the preset center point of the vehicle body meet the parking space occupancy conditions of the target parking frame, then the vehicle of the target vehicle in the multiple frames of the second image collected within the preset time before the first image is acquired. Tracking information, the vehicle tracking information includes the corresponding positioning positions of the vehicle identification of the target vehicle in the multi-frame second images; wherein the target parking frame is the position of any of the parking spots in the camera. The corresponding image area boundary in the shooting screen;

If it is determined according to the vehicle tracking information that the vehicle identification appears in the preset warning frame, it is determined that the target vehicle enters the parking space corresponding to the target parking frame and stops; wherein the warning frame is the entrance to the parking space. The pre-warning area that multiple parking spaces need to pass through is the corresponding image area boundary in the shooting picture of the camera. The early warning area is a roadway area surrounding the multiple parking spaces.
The method according to claim 1, wherein the parking space occupancy condition includes: the coincidence degree between the vehicle detection frame and the target parking frame is greater than a set threshold, and the position of the preset center point of the vehicle body is located at Within the bounding box of the target berth.
The method according to claim 1, wherein determining that the target vehicle drives into the parking space corresponding to the target parking space frame includes:

Obtaining vehicle tracking information of the target vehicle in multiple frames of third images collected within a preset time after the first image;

Carry out vehicle detection processing on the plurality of third frames of images respectively, so as to obtain the corresponding vehicles of the target vehicle in the plurality of frames of third images in combination with the vehicle tracking information of the target vehicle in the plurality of frames of third images. The position of the detection frame and the preset center point of the body;

If the positions of the vehicle detection frame and the preset center point of the vehicle body respectively corresponding to the target vehicle in the multi-frame third image meet the parking space occupancy conditions, it is determined that the target vehicle enters the location corresponding to the target parking space frame. berth.
The method of claim 1, further comprising:

Generate a drive-in record corresponding to the target vehicle. The drive-in record includes the identification of the parking space corresponding to the target parking space frame, the vehicle identification, the drive-in time, and the drive-in video, wherein the drive-in record The video at least includes a video segment sampled from the plurality of frames of the second image and the first image.
The method of claim 1, further comprising:

Perform vehicle detection processing on multiple frames of fourth images collected after the first image respectively;

If it is determined based on the vehicle detection results of the fourth multi-frame image that the target vehicle does not meet the parking occupancy conditions of the target parking frame, then obtain the vehicle tracking information of the target vehicle in the fourth multi-frame image;

If it is determined that the vehicle identification appears in the warning frame according to the vehicle tracking information of the target vehicle in the fourth multi-frame image, then it is determined that the target vehicle leaves the parking space corresponding to the target parking space frame.
The method of claim 5, further comprising:

Generate a drive-out record corresponding to the target vehicle. The drive-out record includes the identity of the parking space corresponding to the target parking space frame, the vehicle identity, the drive-out time, and the drive-out video, wherein the drive-out record The video at least includes video segments sampled from the plurality of frames of fourth images.
The method according to any one of claims 1 to 6, characterized in that the preset center point of the vehicle body includes a vehicle chassis center point.
The method according to claim 7, wherein the vehicle detection processing on the first image includes:

Use the pre-trained deep neural network model to perform vehicle detection processing on the first image to obtain the vehicle detection frame and vehicle chassis detection frame corresponding to the target vehicle in the first image. The vehicle detection frame refers to a vehicle detection frame containing Detection frame of complete vehicle;

The position of the center point of the vehicle chassis detection frame is determined as the position of the center point of the vehicle chassis.
The method according to claim 7, wherein the vehicle detection processing on the first image includes:

Use a pre-trained deep neural network model to perform vehicle detection processing on the first image to obtain a vehicle detection frame corresponding to the target vehicle in the first image;

Determine the sub-detection frame with a preset height ratio at the bottom of the vehicle detection frame, where the preset height ratio is set based on the height of the vehicle chassis from the ground;

The position of the top center point of the sub-detection frame is determined as the position of the center point of the vehicle chassis.
The method according to any one of claims 1 to 6, characterized in that the vehicle identification includes a license plate number and a tracking number, the tracking number is assigned during vehicle tracking processing of the image, and the same vehicle has the same Tracking Number.
An electronic device, characterized by comprising: a memory, a processor, and a communication interface; wherein executable code is stored on the memory, and when the executable code is executed by the processor, the processor The method for determining vehicle entry and exit from a parking space according to any one of claims 1 to 10 is performed.
A non-transitory machine-readable storage medium, characterized in that executable code is stored on the non-transitory machine-readable storage medium. When the executable code is executed by a processor of an electronic device, the The processor executes the method for determining the entry and exit of a vehicle into a parking space as claimed in any one of claims 1 to 10.
A method for determining the entry and exit of vehicles into roadside parking spaces, which is characterized by including:

Obtaining the first image collected by a roadside camera, the roadside camera includes a camera arranged on the same side of the roadside parking space, and the roadside camera covers multiple roadside parking spaces;

Perform vehicle detection processing on the first image to obtain the vehicle detection frame corresponding to the target vehicle in the first image and the position of the preset center point of the vehicle body;

If it is determined that the positions of the vehicle detection frame and the preset center point of the vehicle body meet the parking space occupancy conditions of the target parking frame, then the vehicle of the target vehicle in the multiple frames of the second image collected within the preset time before the first image is acquired. Tracking information, the vehicle tracking information includes the corresponding positioning positions of the vehicle identification of the target vehicle in the multiple frames of second images; wherein the target parking frame is the location of any of the roadside parking spaces in the second image. The corresponding image area boundary in the picture taken by the roadside camera;

If it is determined according to the vehicle tracking information that the vehicle identification appears in the preset warning frame, it is determined that the target vehicle drives into the roadside parking space corresponding to the target parking frame and stops; wherein the warning frame is the entrance and exit The warning area that the multiple roadside parking spaces need to pass through is the corresponding image area boundary in the picture captured by the roadside camera. The warning area is a roadway area surrounding the multiple roadside parking spaces.