Vehicle parking safety intelligent monitoring method based on image recognition

Abstract

The invention discloses a vehicle parking safety intelligent monitoring method based on image recognition, and relates to the field of image recognition. The method comprises the following steps: acquiring a street video image, setting a video image processing model, eliminating and sorting the street video image through the video image processing model, and further acquiring a parking space image and a parking vehicle image; setting a one-way transmission chain, taking parking spaces and parking vehicles as central data nodes, performing connection through the one-way transmission chain to generate a parking space real-time data network and a parking vehicle real-time data network, and further generating a vehicle parking management model; processing the parking space image based on a vehicle parking management model to obtain an abnormal parking space image, and performing safety management on the parking space according to parking space image information data; acquiring an abnormally parked vehicle according to the parked vehicle real-time data network, and performing safety management on the abnormally parked vehicle according to the parked vehicle image information data; according to the invention, the normalization of vehicle parking and the convenience of safety monitoring are improved.

Description

Technical Field

[0001] The present invention relates to the field of image recognition, and specifically to an intelligent monitoring method for vehicle parking safety based on image recognition. Background Art

[0002] Image recognition refers to the technology of using a computer to process, analyze, and understand images to identify various different patterns of targets and objects. It is a practical application of deep learning algorithms. At the present stage, image recognition technology is generally divided into face recognition and object recognition. Image recognition is an important field of artificial intelligence. The recognition of objects mainly refers to the perception and understanding of objects and environments in the three-dimensional world, belonging to the category of advanced computer vision. It is a research direction that combines digital image processing and recognition with disciplines such as artificial intelligence and systems science, and its research results are widely applied to various industrial and exploration robots. Vehicle parking is everywhere in modern life, not only in parking lots but also in various parking spaces, such as street parking spaces, residential parking spaces, park parking spaces, etc. However, due to the irregular parking of vehicles in parking spaces, parking safety problems occur, for example, random parking on the street leads to road congestion and safety problems, and it is impossible to monitor the parking of street vehicles in real time. Manual on-site control of parked vehicles is required, resulting in the inability to manage parked vehicles in a timely manner. To solve the above problems, image recognition is used to monitor vehicle parking. Therefore, an intelligent monitoring method for vehicle parking safety based on image recognition is provided. Summary of the Invention

[0003] To solve the above technical problems, the present invention provides an intelligent monitoring method for vehicle parking safety based on image recognition. The object of the present invention can be achieved by the following technical solutions: An intelligent monitoring method for vehicle parking safety based on image recognition, the method includes the following steps: Step S1: Obtain street video images, set a video image processing model, and eliminate and sort the street video images through the video image processing model to obtain parking space images and parked vehicle images. Step S2: Set a unidirectional transmission chain, using the parking space and the parked vehicle as the central data nodes respectively, for storing parking space image information data and parked vehicle image information data, and connecting them through the unidirectional transmission chain to generate a real-time data network for parking spaces and a real-time data network for parked vehicles, and then generate a vehicle parking management model. Step S3: Process the parking space images based on the vehicle parking management model to obtain abnormal parking space images, and perform safety management on the abnormal parking space images according to the corresponding parking space image information data. Step S4: Obtain abnormal parked vehicles according to the real-time data network of parked vehicles, and conduct safety management on the abnormal parked vehicles based on the image information data of parked vehicles.

[0004] Further, the setting process of the video image processing model includes: Set a specific acquisition time point, and obtain street video images in real time through street cameras according to the specific acquisition time point; set a video image processing model, and the video image processing model includes an elimination unit and a sorting unit; Obtain a number of vehicle models and store them to generate a vehicle model library for wireless communication connection to the elimination unit; The elimination unit is used to perform image elimination on the street video images according to the vehicle model library to obtain street vehicle video images; The sorting unit is used to sort the street vehicle video images to obtain parking space images and parked vehicle images.

[0005] Further, the acquisition process of the parking space images and parked vehicle images includes: Send the street video images to the video image processing model. The elimination unit marks and retains the vehicles corresponding to the vehicle model library in the street video images, and eliminates the remaining unmarked ones, thereby generating street vehicle video images and sending them to the sorting unit; Pre-store the street parking space distribution map in the sorting unit, obtain the vehicles corresponding to the street parking space distribution map in the street vehicle video images, and intercept the corresponding street vehicle video images according to the street parking space distribution map to obtain parking space images, and then intercept the remaining vehicles in the street vehicle video images to obtain parked vehicle images; Respectively obtain the shooting times corresponding to the parking images and parked vehicle images and store them.

[0006] Further, the generation process of the parking space real-time data network and the parked vehicle real-time data network includes: The unidirectional transmission chain includes an update chain and a warning chain; the update chain is used to transmit information data to the central data node for update; the warning chain is used to transmit information data to the terminal for warning, and the terminal includes the vehicle owner's mobile terminal and the management terminal; Generate a parking space central data node and a parked vehicle central data node for the parking space and the parked vehicle respectively and use them as the central data nodes; Generate a parking space image node and a parked vehicle image node for the parking space images and parked vehicle images respectively, and connect the parking space image node and the parked vehicle image node to the parking space central data node and the parked vehicle central data node respectively through the update chain, thereby generating a parking space real-time data network and a parked vehicle real-time data network respectively.

[0007] Further, the generation process of the vehicle parking management model includes: Obtain the parking space image information data and the parked vehicle image information data corresponding to the parking space image and the parked vehicle image respectively, and send them to the corresponding parking space central data node and parked vehicle central data node for connection and storage; The parking space image information data includes license plate number, contact information, parking space code, and shooting time; the parked vehicle image information data includes license plate number, contact information, parking location, and shooting time; Connect the parking space real-time data network and the parked vehicle real-time data network through a data link to generate a vehicle parking management model; Further, the process of obtaining the abnormal parking space image includes: Perform edge processing on the parking space image according to the street parking space distribution map to obtain a standard parking space image. Judge the vehicle integrity of the standard parking space image according to the vehicle model. If there is a complete vehicle model in the standard parking space image, the corresponding vehicle is in a complete state, and a safe parking space image is generated. Otherwise, the corresponding vehicle is in an incomplete state, and an abnormal parking space image is generated.

[0008] Further, the process of performing safety management on the parking space includes: Obtain the parking space image information data corresponding to the abnormal parking space image, and obtain the contact information according to the parking space image information data. Then, send the license plate number and the parking space code to the vehicle owner terminal through the warning chain according to the contact information, and send the deadline of the vehicle specification processing time for warning; Obtain the street video image again according to the deadline of the vehicle specification processing time, and then obtain the parking space image corresponding to the abnormal parking space image. Judge the integrity of the abnormal parking space image again. If it is in a complete state, generate a safe parking space image and remove the corresponding parking space image node; if it is in an incomplete state, send the parking space image information data to the management terminal through the warning chain. After the management terminal receives the corresponding parking space image information data, obtain the vehicle corresponding to the license plate number according to the parking space code and give a penalty warning to the vehicle; if there is no vehicle model in the abnormal parking space image, generate a vacant parking space image, and remove and update the corresponding parking space image node from the parking space real-time data network.

[0009] Further, the process of performing safety management on the abnormal parked vehicle includes: Mark all the parking vehicle images corresponding to the parking vehicle image nodes in the real-time parking vehicle data network as abnormal parking vehicles, obtain the corresponding parking vehicle image information data according to the parking vehicle central data node, send a warning to the vehicle owner's mobile terminal through the warning chain according to the contact information, and send the deadline for the vehicle's standardized handling time for warning; Similarly, obtain the parking vehicle image corresponding to the abnormal parking vehicle again according to the deadline for the vehicle's standardized handling time and make a judgment. If the vehicle has not parked in a standardized manner, send the parking vehicle image information data to the management terminal through the warning chain for penalty management; if the vehicle has parked in a standardized manner, remove the corresponding parking vehicle image node and update the real-time parking vehicle data network.

[0010] Compared with the prior art, the beneficial effects of the present invention are: 1. Obtain street video images, set up a video image processing model, eliminate and sort the street video images through the video image processing model, and then obtain parking space images and parking vehicle images; set up a one-way transmission chain, use the parking space and the parking vehicle as central data nodes respectively, and connect them through the one-way transmission chain to generate a real-time parking space data network and a real-time parking vehicle data network, and then generate a vehicle parking management model, which solves the convenience of vehicle parking; 2. Judge the integrity of the parking space image, obtain abnormal parking space images, then obtain the corresponding parking space image information data through the central data node, and send a warning to the vehicle owner through the warning chain. Obtain the abnormal parking space images again according to the deadline for the vehicle's standardized handling time, and judge the integrity again, and then judge whether the vehicle corresponding to the abnormal parking space image has been standardized, and update the corresponding real-time parking space data network; through the real-time parking space data network, the abnormal parking space images can be monitored more clearly, and warning notifications can be managed through the warning chain, improving the standardization of street vehicle parking. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required to be used in the embodiments. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those of ordinary skill in the art, other drawings can also be obtained according to these drawings.

[0012] Figure 1 It is a flowchart of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0013] To more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the accompanying drawings required for use in the embodiments. Obviously, the accompanying drawings in the following description are only some embodiments recorded in the present invention. For those of ordinary skill in the art, other accompanying drawings can also be obtained based on these drawings. Embodiment

[0014] As Figure 1 shown, a vehicle parking safety intelligent monitoring method based on image recognition, the method includes the following steps: Step S1: Obtain street video images, and perform elimination and sorting to obtain parking space images and parked vehicle images; Step S2: Set up a unidirectional transmission chain, using the parking space and the parked vehicle as the central data nodes respectively, for storing parking space image information data and parked vehicle image information data, and connecting them through the unidirectional transmission chain to generate a real-time data network for parking spaces and a real-time data network for parked vehicles, and then generate a vehicle parking management model; Step S3: Process the parking space images based on the vehicle parking management model to obtain abnormal parking space images, and perform safety management on the abnormal parking space images according to the corresponding parking space image information data; Step S4: Obtain abnormal parked vehicles according to the real-time data network of parked vehicles, and perform safety management on the abnormal parked vehicles according to the parked vehicle image information data; Embodiment

[0015] This embodiment is a further limitation of Embodiment 1, and the step S1 is implemented through the following process: Set a specific acquisition time point, and obtain street video images in real time through street cameras according to the specific acquisition time point; set a video image processing model, and the video image processing model includes an elimination unit and a sorting unit; Obtain a number of vehicle models and store them to generate a vehicle model library, which is wirelessly connected to the elimination unit; The elimination unit is used to perform image elimination on the street video images according to the vehicle model library to obtain street vehicle video images; The sorting unit is used to sort the street vehicle video images to obtain parking space images and parked vehicle images; It should be further noted that in the specific implementation process, if the street video images obtained on the street are the street video images obtained throughout the day, it is impossible to determine whether the vehicle is in a parked state or a traffic jam state during the peak commuting hours. Therefore, a characteristic acquisition time point is set, and the characteristic acquisition time point includes but is not limited to morning, noon, evening, night, etc.; in this way, the obtained street video images can better judge the parked state of the vehicle. Send the street video image to the video image processing model. Mark and retain the vehicles corresponding to the vehicle model library in the street video image through the elimination unit, and eliminate the remaining unmarked ones, and then generate a street vehicle video image and send it to the sorting unit; It should be further noted that in the specific implementation process, the vehicle model is used to represent the form of the vehicle, including but not limited to cars, large trucks, convertible cars, etc.; eliminating the miscellaneous people in the street video according to the vehicle model library improves the clarity and processing convenience of the street video image; Pre-store the street parking space distribution map in the sorting unit, obtain the vehicles corresponding to the street parking space distribution map in the street vehicle video image, and intercept the corresponding street vehicle video image according to the street parking space distribution map to obtain the parking space image, and then intercept the remaining vehicles in the street vehicle video image to obtain the parked vehicle image; Obtain and store the shooting times corresponding to the parking image and the parked vehicle image respectively; It should be further noted that in the specific implementation process, the vehicle images on the street are divided into two categories, one is the parking space image and the parked vehicle image, which can better analyze the vehicles on the street, and then conduct safety management on vehicle parking; Embodiment

[0016] This embodiment is a further limitation of Embodiment 1, and the step S2 is implemented through the following process: The unidirectional transmission chain includes an update chain and a warning chain; the update chain is used to transmit information data to the central data node for update; the warning chain is used to transmit information data to the terminal for warning, and the terminal includes the vehicle owner's mobile terminal and the management terminal; Generate a parking space central data node and a parked vehicle central data node for the parking space and the parked vehicle respectively and use them as the central data nodes; Generate a parking space image node and a parked vehicle image node for the parking space image and the parked vehicle image respectively, and connect the parking space image node and the parked vehicle image node with the parking space central data node and the parked vehicle central data node respectively through the update chain, and then generate a parking space real-time data network and a parked vehicle real-time data network respectively; Obtain the parking space image information data and the parked vehicle image information data corresponding to the parking space image and the parked vehicle image respectively and send them to the corresponding parking space central data node and parked vehicle central data node for connection and storage; Further, the parking space image information data includes the license plate number, contact information, parking space code and shooting time; the parked vehicle image information data includes the license plate number, contact information, parking location and shooting time; Connect the real-time data network of parking spaces and the real-time data network of parked vehicles through a data link to generate a vehicle parking management model; It should be further noted that in the specific implementation process, by setting up a one-way transmission link for the update chain and the warning chain, information data can be better saved to prevent the leakage of information data; generating corresponding data nodes can better process the data; Embodiment

[0017] This embodiment is a further limitation of Embodiment 1, and the step S3 is implemented through the following process: Perform edge processing on the parking space image according to the street parking space distribution map to obtain a standard parking space image. Judge the integrity of the vehicle in the standard parking space image according to the vehicle model. If there is a complete vehicle model in the standard parking space image, the corresponding vehicle is in a complete state, and a safe parking space image is generated. Otherwise, the corresponding vehicle is in an incomplete state, and an abnormal parking space image is generated; Obtain the parking space image information data corresponding to the abnormal parking space image, and obtain the contact information according to the parking space image information data. Then, send the license plate number and the parking space code to the vehicle owner's terminal through the warning chain, and send the deadline for the vehicle specification processing time for warning; Send the vehicle specification processing time to the corresponding parking space image node for storage; Obtain the street video image again according to the deadline for the vehicle specification processing time, and then obtain the parking space image corresponding to the abnormal parking space image. Judge the integrity of the abnormal parking space image again. If it is in a complete state, generate a safe parking space image and remove the corresponding parking space image node; if it is in an incomplete state, send the parking space image information data to the management terminal through the warning chain. When the management terminal receives the corresponding parking space image information data, obtain the vehicle corresponding to the license plate number according to the parking space code and give a penalty warning to the vehicle; if there is no vehicle model in the abnormal parking space image, generate an empty parking space image, and remove and update the corresponding parking space image node from the real-time data network of parking spaces; It should be further noted that in the specific implementation process, by judging the integrity of the parking space image, obtaining the abnormal parking space image, then obtaining the corresponding parking space image information data through the central data node, and sending a warning to the vehicle owner through the warning chain. Obtain the abnormal parking space image again according to the deadline for the vehicle specification processing time, and judge the integrity again, and then judge whether the vehicle corresponding to the abnormal parking space image is processed according to the specification, and update the corresponding real-time data network of parking spaces; through the real-time data network of parking spaces, the abnormal parking space image can be monitored more clearly, and warning notifications can be sent through the warning chain for management, improving the standardization of street vehicle parking; Embodiment

[0018] This embodiment is a further limitation of Embodiment 1, and the step S4 is implemented through the following process: All the parking vehicle images corresponding to the parking vehicle image nodes in the real-time data network of parked vehicles are marked as abnormally parked vehicles. The parking vehicle image information data is obtained according to the parking vehicle center data node, and is sent to the mobile terminal of the vehicle owner through the warning chain for warning, and the deadline for the vehicle specification processing time is sent for warning; Similarly, according to the deadline for the vehicle specification processing time, the parking vehicle images corresponding to the abnormally parked vehicles are obtained again and judged. If the vehicle has not parked in a standardized manner, the parking vehicle image information data is sent to the management terminal through the warning chain for penalty management; if the vehicle has parked in a standardized manner, the corresponding parking vehicle image node is removed and the real-time data network of parked vehicles is updated.

[0019] It should be further noted that the abnormally parked vehicles on the street are classified into abnormally parked vehicles in the parking space and abnormally parked vehicles not in the parking space. The two types of abnormally parked vehicles are classified and processed, and a warning chain is set up. The abnormally parked vehicles are monitored for classification levels through the warning chain, which improves the accuracy of monitoring; The features and exemplary embodiments of various aspects of the present application will be described in detail above. In order to make the purpose, technical solution and advantages of the present application clearer, the present application will be further described in detail below in combination with the accompanying drawings and specific embodiments; it should be understood that the specific embodiments described here are only intended to explain the present application, rather than limit the present application; for those skilled in the art, the present application can be implemented without some of these specific details. The above description of the embodiments is only to provide a better understanding of the present application by showing examples of the present application.

[0020] The above embodiments are only used to illustrate the technical method of the present invention and not to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical method of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A vehicle parking safety intelligent monitoring method based on image recognition, characterized in that: The method comprises the following steps: Step S1: obtaining a street video image, setting a video image processing model, eliminating and sorting the street video image through the video image processing model, and then obtaining a parking space image and a parked vehicle image; Step S2: Setting a one-way transmission chain, setting parking spaces and parked vehicles as central data nodes respectively, for storing parking space image information data and parking vehicle image information data, connecting through the one-way transmission chain to generate a parking space real-time data network and a parking vehicle real-time data network, and then generating a vehicle parking management model; Step S3: Processing the parking space image based on the vehicle parking management model, obtaining abnormal parking space images, and performing safety management on the abnormal parking space images according to corresponding parking space image information data; Step S4: Acquire abnormal parked vehicles based on the real-time data network of parked vehicles, and perform safety management on the abnormal parked vehicles based on the image information data of the parked vehicles.

2. The method for intelligent monitoring of vehicle parking safety based on image recognition according to claim 1, characterized in that: The setting process of the video image processing model includes: Setting a specific acquisition time point, and acquiring street video images in real time through a street camera according to the specific acquisition time point; setting a video image processing model, wherein the video image processing model includes an elimination unit and a sorting unit; Acquire several vehicle models and store them to generate a vehicle model library wireless communication connection elimination unit; The elimination unit is used to perform image elimination on the street video image according to the vehicle model library to obtain the street vehicle video image; The sorting unit is used to sort the street vehicle video images to obtain parking space images and parked vehicle images.

3. The method for intelligent monitoring of vehicle parking safety based on image recognition according to claim 2 is characterized in that: The process of acquiring the parking space image and the parking vehicle image includes: The street video image is sent to the video image processing model, and the vehicles corresponding to the vehicle model library in the street video image are marked and retained by the elimination unit, and the remaining unmarked ones are eliminated, thereby generating a street vehicle video image and sending it to the sorting unit; The street parking space distribution map is pre-stored in the sorting unit, and the vehicle corresponding to the street parking space distribution map in the street vehicle video image is obtained, and the street parking space distribution map is intercepted from the corresponding street vehicle video image to obtain a parking space image, and then the remaining vehicles in the street vehicle video image are intercepted to obtain a parking vehicle image; The shooting times corresponding to the parking image and the parking vehicle image are respectively obtained and stored.

4. The method for intelligent monitoring of vehicle parking safety based on image recognition according to claim 3 is characterized in that: The generation process of the parking space real-time data network and the parking vehicle real-time data network includes: The one-way transmission chain includes an update chain and an early warning chain; the update chain is used to transmit information data to a central data node for storage; the early warning chain is used to transmit information data to a terminal for early warning; the terminal includes a vehicle owner mobile terminal and a management terminal; The parking spaces and the parking vehicles are respectively generated as parking space central data nodes and parking vehicle central data nodes and used as central data nodes; The parking space image and the parking vehicle image are respectively generated into a parking space image node and a parking vehicle image node, and the parking space image node and the parking vehicle image node are respectively connected to the parking space central data node and the parking vehicle central data node through an update chain, thereby generating a parking space real-time data network and a parking vehicle real-time data network respectively.

5. The method for intelligent monitoring of vehicle parking safety based on image recognition according to claim 4 is characterized in that: The generation process of the vehicle parking management model includes: Obtaining parking space image information data and parking vehicle image information data corresponding to the parking space image and the parking vehicle image respectively and sending them to the corresponding parking space central data node and parking vehicle central data node for connection and storage; The parking space image information data includes the license plate number, contact information, parking space code and shooting time; the parking vehicle image information data includes the license plate number, contact information, parking location and shooting time; The parking space real-time data network and the parking vehicle real-time data network are connected through a data link to generate a vehicle parking management model.

6. The method for intelligent monitoring of vehicle parking safety based on image recognition according to claim 5, characterized in that: The process of acquiring the abnormal parking space image includes: The parking space image is edge processed according to the street parking space distribution map to obtain the standard parking space image, and the vehicle integrity of the standard parking space image is judged according to the vehicle model. If the standard parking space image has a complete vehicle model, the corresponding vehicle is in a complete state and a safe parking space image is generated. Otherwise, the corresponding vehicle is in an incomplete state and an abnormal parking space image is generated.

7. The method for intelligent monitoring of vehicle parking safety based on image recognition according to claim 6, characterized in that: The process of safely managing parking spaces includes: Obtain parking space image information data corresponding to the abnormal parking space image, and obtain contact information based on the parking space image information data, and then send the license plate number and parking space code to the vehicle owner terminal through the early warning chain according to the contact information, and send the vehicle standard processing time deadline for early warning; According to the deadline of vehicle specification processing time, the street video image is acquired again, and then the parking space image corresponding to the abnormal parking space image is acquired, and the integrity of the abnormal parking space image is judged again. If it is in a complete state, a safe parking space image is generated, and the corresponding parking space image node is removed; if it is in an incomplete state, the parking space image information data is sent to the management terminal through the early warning chain. The management terminal receives the corresponding parking space image information data, and obtains the vehicle corresponding to the license plate number according to the parking space code, and issues a penalty warning to the vehicle; if there is no vehicle model in the abnormal parking space image, an empty parking space image is generated, and the corresponding parking space image node is removed and updated from the parking space real-time data network.

8. The method for intelligent monitoring of vehicle parking safety based on image recognition according to claim 5, characterized in that: The process of safely managing abnormally parked vehicles includes: Mark all the parking vehicle images corresponding to the parking vehicle image nodes in the parking vehicle real-time data network as abnormal parking vehicles, obtain the corresponding parking vehicle image information data according to the parking vehicle center data node, send it to the vehicle owner's mobile terminal through the early warning chain according to the contact information for early warning, and send the vehicle standard processing time deadline for early warning; Similarly, according to the deadline for vehicle standard processing, the parking vehicle image corresponding to the abnormally parked vehicle is obtained again and judged. If the standard parking is not performed, the parking vehicle image information data is sent to the management terminal through the early warning chain for penalty management; if the standard parking is performed, the corresponding parking vehicle image node is removed and the parking vehicle real-time data network is updated.

Images