CN112419778A - Parking lot vacancy detection and optimal path guiding device based on image processing and vacancy detection and optimal path guiding method thereof - Google Patents
Parking lot vacancy detection and optimal path guiding device based on image processing and vacancy detection and optimal path guiding method thereof Download PDFInfo
- Publication number
- CN112419778A CN112419778A CN202011235308.6A CN202011235308A CN112419778A CN 112419778 A CN112419778 A CN 112419778A CN 202011235308 A CN202011235308 A CN 202011235308A CN 112419778 A CN112419778 A CN 112419778A
- Authority
- CN
- China
- Prior art keywords
- module
- information
- parking lot
- detection
- parking
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/14—Traffic control systems for road vehicles indicating individual free spaces in parking areas
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/14—Traffic control systems for road vehicles indicating individual free spaces in parking areas
- G08G1/145—Traffic control systems for road vehicles indicating individual free spaces in parking areas where the indication depends on the parking areas
- G08G1/148—Management of a network of parking areas
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N7/00—Television systems
- H04N7/18—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast
- H04N7/181—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast for receiving images from a plurality of remote sources
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Traffic Control Systems (AREA)
Abstract
A parking lot vacancy detection and optimal path guiding device based on image processing and a vacancy detection and optimal path guiding method thereof comprise a parking lot vacancy detection device, an optimal path guiding device, a video detection early warning device, a database and management system and an information display device; the database and management system is used for carrying out coordination management and control on information interaction and transmission among the parking space detection device, the optimal path guiding device element, the video detection early warning device element and the information display device; the invention utilizes video monitoring image processing, database data processing and computer vision optimization algorithm, has low installation cost and simple layout, and can quickly and efficiently realize the vacancy detection and the optimal path guidance of the parking lot.
Description
Technical Field
The invention relates to the field of intelligent detection and management of parking lots, in particular to a vacancy detection and optimal path guiding device based on video image processing and a vacancy detection and optimal path guiding method thereof.
Background
In recent years, dynamic balancing and static management of automobiles have become the focus of research for many experts and scholars. The intelligent management system for the parking lot urgently needs to improve the management level of the parking lot by adopting a more automatic, more scientific, more standard, effective, faster, safe and convenient mode. At present, with the development of various scientific technologies, the intelligent research of a parking lot is mainly used for newly building the parking lot, and the intelligent reconstruction research of an old parking lot is relatively less; the automatic detection is mainly used for reasonably charging vehicles in the vehicle entrance and exit link, the intelligent detection and the standardized utilization research of parking spaces in a parking lot are less, and the matching of a single parking space and the parked vehicles cannot be realized. Simultaneously, current parking area topography is complicated, and the subregion is loaded down with trivial details, and the vehicle drives into and seeks the parking stall and need winding many times usually, brings very big inconvenience for the driver when extravagant the energy, and the solution method of present parking area route guide is the shortest path guide mostly, for example: patent document CN111243327A discloses a vehicle guidance method based on computer monitoring, which is to locate the current position of a vehicle, screen out a target parking space closest to the current position from a set of free parking spaces, and plan a shortest path to the target parking space to guide the vehicle. However, other route guidance of people with special parking requirements is not considered, and meanwhile, the single guidance mode is poor in overall harmony and not beneficial to reasonable planning and utilization of large parking lot partitions. Secondly, the resource allocation of the existing parking lot is nervous, the parking spaces are densely distributed, the events of scratching and interference caused by the nonstandard parking of a driver are easily induced to a certain extent, and the development trend of the intelligent parking lot prompts that modern technology is utilized to warn and manage the events.
Disclosure of Invention
Based on the defects in the prior art, the invention provides a parking lot vacancy detection and optimal path guiding device based on image processing and a vacancy detection and optimal path guiding method thereof. The invention utilizes video monitoring image processing, database data processing and computer vision optimization algorithm, has low installation cost and simple layout, and can quickly and efficiently realize the vacancy detection and the optimal path guidance of the parking lot. In order to achieve the purpose, the invention adopts the following technical scheme:
a parking lot vacancy detection and optimal path guiding device based on image processing comprises a parking lot detection device, an optimal path guiding device, a video detection early warning device, a database and management system and an information display device;
the database and management system is used for carrying out coordination management and control on information interaction and transmission among the parking space detection device, the optimal path guiding device element, the video detection early warning device element and the information display device;
the parking space detection device is used for identifying the existing parking condition and the number and positions of the empty parking spaces in the parking lot and simultaneously outputting the parking conditions and the numbers and positions to the database and the management system for real-time scheduling;
the optimal path guiding device is used for carrying out region division and optimal path guiding on vehicles entering the parking lot by using a data processing and deep learning algorithm so as to ensure that the optimal path is matched for the entering vehicles and be beneficial to reasonable planning and utilization of large parking lot partitions;
the video detection early warning device is used for carrying out safety early warning on a dynamic vehicle target in a parking lot so as to avoid the obstruction and the scraping of other vehicles caused by improper warehousing directions;
the information display device is used for receiving the path guiding information and the early warning information output by the database and the management system, and the path guiding information and the early warning information are output through a large screen of the parking lot, a guide rail indicator lamp of a ceiling and a loudspeaker to remind driving.
The parking space detection device comprises an image acquisition module, a first computer image processing module and a first output module; the image acquisition module comprises camera equipment, gateway equipment and transmission equipment connected with the camera equipment and is used for carrying out video monitoring on the parking lot by setting the frame number; the first computer image processing module relates to a real-time video target detection algorithm, dynamic target identification, key frame comparison and camera image receiving for unification processing; the first output module is used for outputting the parking space detection result to the database and management system.
The parking space detection device comprises the following specific working steps:
s1: the camera equipment in the parking lot is connected with the gateway equipment, tracks and identifies moving vehicles and personnel, and sends digital signals of parking space state information and image information to the computer background system;
s2: the computer dynamic target recognition algorithm is designed based on OpenCV, and is used for detecting a moving target under a dynamic background, wherein the three steps are motion estimation, motion compensation and target detection;
1) and (3) motion estimation: and (3) finding a unique matching point of each characteristic point of the reference image in the image by adopting a characteristic point matching algorithm, calculating a displacement vector through the position change of the point between two frames, and substituting all motion information into a motion model to calculate the global motion parameter of the background. The reference image feature point is first matched with the current feature point by taking the current feature point as the center and taking 4 as the step, and the center position of the feature point with the best match is obtained. Secondly, changing the stride into 2 and 1 to repeat in sequence on the basis;
2) and motion compensation: affine transformation model parameters are used for replacing and estimating motion vectors of all pixel points, a least square method is adopted for solving known characteristic points, and background movement caused by scanning and shaking of the camera is eliminated by means of superposition of a reverse motion compensation frame image and a reference frame image;
3) and detecting the target. Performing frame difference on the stable sequence of the background image to obtain a moving target area, and realizing dynamic target detection;
s3: the dynamic target detection transmits the identified empty parking space increase and occupation conditions caused by the vehicle warehouse-out and warehouse-in to the key frame extraction module; the key frame extraction module is used for identifying and comparing a dynamic difference frame and a front static frame and a rear static frame in the video acquisition and dynamic identification processes, further confirming the use condition of an empty parking space, and meanwhile outputting a comparison and verification result to the output module;
s4: the first output module outputs the detection result to the database and management module, and updates the database information.
The optimal path guiding device comprises an entrance node license plate information acquisition module, a database information matching module and a second output module; the entrance node license plate information acquisition module comprises entrance camera equipment, gateway equipment and transmission equipment connected with the camera equipment and is used for identifying license plate numbers of vehicles driven into the entrance of the parking lot; the database information matching module comprises a data processing module and a deep learning algorithm module, and learns the optimal parking area and path by using the historical data of the vehicle; and the second output module is used for outputting the data matching result to the database and management system and the information display device, and is used for updating the database and displaying the large-screen and ceiling guide rail indicator lamp of the parking lot.
The specific working steps of the optimal path guiding device are as follows:
s1: the camera equipment at the entrance of the parking lot is connected with the gateway equipment, identifies the license plate number of a vehicle entering the entrance of the parking lot, and sends the digital signal of the license plate information to the computer background system;
s2: the computer receives the license plate information from the camera equipment and inputs the input signal into the database;
s3: the data processing comprises fuzzy information deletion, data fitting processing, data reduction, condition query, time processing and database establishment;
s4: if the license plate information is a registered vehicle, the deep learning module fuses the registration information with historical data of a target vehicle and the existing empty vehicle data of the parking lot, outputs the data to a space-time convolution layer, performs convolutional neural network learning to extract parking characteristics, and outputs an optimal parking space and a parking path according to the empty space condition;
s5: if the license plate information is unregistered but historical driving data exists, the deep learning module firstly calls database data to find historical information of the vehicle for parking in the parking lot, wherein the historical information comprises driving frequency, driving time, parking time and the like; the deep learning module fuses historical data of a target vehicle and the existing empty vehicle data of the parking lot, outputs the data to a space-time convolution layer, performs convolutional neural network learning to extract parking characteristics, and outputs an optimal parking space and a parking path by combining with the empty parking space condition;
s6: if the license plate information is the unregistered vehicle and no historical driving data exists, the deep learning module directly utilizes the existing empty vehicle bit data of the parking lot to output shortest path planning for the driving vehicle so that a driver can find an empty vehicle space as soon as possible;
s7: the parking space allocation result is transmitted to a second output module; and the second output module outputs the data matching result to the database and management system and the information display device for updating the database and displaying the large-screen and ceiling guide rail indicator lamp of the parking lot.
The video detection early warning device comprises a video monitoring module, a second computer image processing module and a third output module; the video monitoring module comprises camera equipment, gateway equipment and transmission equipment connected with the camera equipment and is used for carrying out video monitoring on the vehicle parking process in the parking lot; the second computer image processing module relates to a real-time video target detection algorithm and a key frame comparison technology, receives parking space sideline information detected by a camera, and identifies vehicles which are too close to the sideline and have potential safety hazards; and the third output module is used for outputting the safety early warning result to the database and management system and the information display device, and is used for inputting the management system and switching on and off the loudspeaker in the parking lot.
The video detection early warning device comprises the following specific working steps:
s1: the camera equipment in the parking lot is connected with the gateway equipment, tracks and identifies moving vehicles and personnel, and sends digital signals of parking state information and image information to the computer background system;
s2: the computer receives the monitoring information from the camera equipment and positions and tracks the vehicle which is going to enter or exit the garage by combining the dynamic target recognition algorithm of the computer image processing module of the parking space detection device.
S3: and in the real-time video target detection, a Canny edge detection algorithm under a static background is adopted for edge detection.
The Canny edge detection algorithm comprises three steps of filtering processing, edge enhancement and thresholding;
1) and carrying out gray level processing on the input image, and carrying out weighted average according to sampling values of all channels of the image.
2) The filtering process adopts Gaussian filtering, a group of normalized Gaussian kernels are generated by adopting a discretized Gaussian function, and then each point of the image gray matrix is subjected to weighted summation based on the Gaussian kernels, so that the performance of the edge detector is improved.
3) In the edge enhancement process, points with significant changes in the intensity values of the neighborhood of the gray points of the image are highlighted by calculating the gradient amplitude, and the change values of the neighborhood intensity of each point of the image are determined.
4) And discarding points with larger gradient values but not edge points in the neighborhood by adopting a thresholding method.
S4: calculating the deviation position of the vehicle relative to the vehicle line; if the vehicle is too close to the sideline, a loudspeaker of the information display unit is triggered to emit warning sound to warn; if no sideline is too close to the problem, triggering is not carried out;
s5: the video detection early warning result is transmitted to the third output module, and the third output module outputs the safety early warning result to the database and management system and the information display device for inputting the management system and opening and closing the loudspeaker in the parking lot.
The database and management system comprises a computer background, a first controller module, a mobile communication module, a motherboard bus and a database system; the first controller module and the mobile communication module are in communication connection and information interaction with a motherboard bus through a computer background;
the first controller module is used for acquiring, analyzing, processing, coordinating and controlling information of each functional unit in the parking lot vacancy detection and optimal path guiding device; the mobile communication module is used for transmitting parking state information, running state information and running log information of the parking lot vacancy detection and optimal path guiding device to a remote management center for storage and emergency intervention management of emergency personnel; the mother board bus is used for carrying out communication connection and information interaction with each functional unit in the parking lot vacancy detection and optimal path guiding device; the database system is used for updating and inputting the data of the information output by each functional unit in the parking lot vacancy detection and optimal path guiding device.
The information display device comprises an LED display module, a loudspeaker module and a second controller module; the LED display module and the loudspeaker module are in communication connection and information interaction with each functional unit in the parking lot vacancy detection and optimal path guiding device through a mother board bus and a controller;
the LED display module comprises a large parking lot screen and a ceiling guide rail indicator lamp, the large parking lot screen is used for displaying the current parking space condition in the current parking lot, and the ceiling guide rail indicator lamp displays the route of the vehicle entering the parking lot according to the calculated optimal path through the color of the lamp; the loudspeaker module is used for carrying out safe voice reminding on the parked vehicle; the second controller module is used for carrying out information acquisition, analysis, processing, coordination and control on an optimal path guiding unit and a video detection early warning unit in the parking lot vacancy detection and optimal path guiding device.
The invention has the beneficial effects that:
the invention utilizes video monitoring image processing, database data processing and computer vision optimization algorithm, has low installation cost and simple layout, and can quickly and efficiently realize the vacancy detection and the optimal path guidance of the parking lot.
Drawings
Fig. 1 is a block diagram of the parking space detection and optimal path guidance apparatus based on image processing according to the present invention.
Fig. 2 is a detailed block diagram of the parking space detection and optimal path guidance device based on image processing according to the present invention.
Fig. 3 is a flowchart of the operation of the parking space detection and optimal path guidance device based on image processing according to the present invention.
Detailed Description
As shown in fig. 1, an image processing-based parking lot vacancy detection and optimal path guiding device includes a parking space detection device 1, an optimal path guiding device 2, a video detection early warning device 3, a database and management system 4, and an information display device 5;
the database and management system 4 is used for performing coordination management and control on information interaction and transmission among the parking space detection device 1, the optimal path guiding device 2, the video detection early warning device 3 and the information display device 5;
the parking space detection device 1 is used for identifying the existing parking condition and the number and positions of empty parking spaces in the parking lot and simultaneously outputting the parking conditions and the number and positions to the database and management system 4 for real-time scheduling;
the optimal path guiding device 2 is used for carrying out region division and optimal path guiding on vehicles entering the parking lot by using a data processing and deep learning algorithm so as to ensure that the optimal path is matched for the vehicles entering the parking lot, and is beneficial to reasonable planning and utilization of large parking lot partitions;
the video detection early warning device 3 is used for carrying out safety early warning on dynamic vehicle targets in the parking lot so as to avoid the obstruction and the scraping of other vehicles caused by improper warehousing directions;
the information display device 5 is used for receiving the path guiding information and the early warning information output by the database and management system 4, and outputting the path guiding information and the early warning information through a large screen of the parking lot, a guide rail indicator lamp of a ceiling and a loudspeaker to remind driving.
As shown in fig. 2, the parking space detecting device 1 includes an image collecting module 11, a first computer image processing module 12 and a first output module 13; the image acquisition module comprises camera equipment, gateway equipment and transmission equipment connected with the camera equipment and is used for carrying out video monitoring on the parking lot by setting the frame number; the first computer image processing module 12 relates to a real-time video target detection algorithm, dynamic target identification, key frame comparison and camera image receiving for unification processing; the first output module 13 is configured to output the parking space detection result to the database and management system 4.
As shown in fig. 3, the parking space detection method of the parking space detection device 1 is as follows:
s1: the camera equipment in the parking lot is connected with the gateway equipment, tracks and identifies moving vehicles and personnel, and sends digital signals of parking space state information and image information to the computer background system;
s2: the computer dynamic target recognition algorithm is designed based on OpenCV, and is used for detecting a moving target under a dynamic background, wherein the three steps are motion estimation, motion compensation and target detection;
1) and (3) motion estimation: and (3) finding a unique matching point of each characteristic point of the reference image in the image by adopting a characteristic point matching algorithm, calculating a displacement vector through the position change of the point between two frames, and substituting all motion information into a motion model to calculate the global motion parameter of the background. The reference image feature point is first matched with the current feature point by taking the current feature point as the center and taking 4 as the step, and the center position of the feature point with the best match is obtained. Secondly, changing the stride into 2 and 1 to repeat in sequence on the basis;
2) and motion compensation: affine transformation model parameters are used for replacing and estimating motion vectors of all pixel points, a least square method is adopted for solving known characteristic points, and background movement caused by scanning and shaking of the camera is eliminated by means of superposition of a reverse motion compensation frame image and a reference frame image;
3) and detecting the target. Performing frame difference on the stable sequence of the background image to obtain a moving target area, and realizing dynamic target detection;
s3: the dynamic target detection transmits the identified empty parking space increase and occupation conditions caused by the vehicle warehouse-out and warehouse-in to the key frame extraction module; the key frame extraction module is used for identifying and comparing a dynamic difference frame and a front static frame and a rear static frame in the video acquisition and dynamic identification processes, further confirming the use condition of an empty parking space, and meanwhile outputting a comparison and verification result to the output module;
s4: the first output module 13 outputs the detection result to the database and management module, and updates the database information.
As shown in fig. 2, the optimal path guiding device 2 includes an entrance node license plate information collecting module 21, a database information matching module 22, and a second output module 23; the entrance node license plate information acquisition module 21 comprises entrance camera equipment, gateway equipment and transmission equipment connected with the camera equipment, and is used for identifying license plate numbers of vehicles entering the entrance of the parking lot; the database information matching module 22 comprises a data processing module and a deep learning algorithm module, and learns the optimal parking area and path by using the historical data of the vehicle; the second output module 23 is configured to output the data matching result to the database and management system 4 and the information display device 5, and is configured to update the database and display the large-screen ceiling guide track indicator lamp of the parking lot.
As shown in fig. 3, the optimal path guidance device 2 specifically operates as follows:
s1: the camera equipment at the entrance of the parking lot is connected with the gateway equipment, identifies the license plate number of a vehicle entering the entrance of the parking lot, and sends the digital signal of the license plate information to the computer background system;
s2: the computer receives the license plate information from the camera equipment and inputs the input signal into the database;
s3: the data processing comprises fuzzy information deletion, data fitting processing, data reduction, condition query, time processing and database establishment;
s4: if the license plate information is a registered vehicle, the deep learning module fuses the registration information with historical data of a target vehicle and the existing empty vehicle data of the parking lot, outputs the data to a space-time convolution layer, performs convolutional neural network learning to extract parking characteristics, and outputs an optimal parking space and a parking path according to the empty space condition;
s5: if the license plate information is unregistered but historical driving data exists, the deep learning module firstly calls database data to find historical information of the vehicle for parking in the parking lot, wherein the historical information comprises driving frequency, driving time, parking time and the like; the deep learning module fuses historical data of a target vehicle and the existing empty vehicle data of the parking lot, outputs the data to a space-time convolution layer, performs convolutional neural network learning to extract parking characteristics, and outputs an optimal parking space and a parking path by combining with the empty parking space condition;
s6: if the license plate information is the unregistered vehicle and no historical driving data exists, the deep learning module directly utilizes the existing empty vehicle bit data of the parking lot to output shortest path planning for the driving vehicle so that a driver can find an empty vehicle space as soon as possible;
s7: the parking space allocation result is transmitted to the second output module 23; the second output module 23 outputs the data matching result to the database and management system 4 and the information display device 5, and is used for updating the database and displaying the large screen and ceiling guide rail indicator lamp of the parking lot.
As shown in fig. 2, the video detection and early warning apparatus 3 includes a video monitoring module 31, a second computer image processing module 32, and a third output module 33; the video monitoring module 31 comprises a camera device, a gateway device and a transmission device connected with the camera device, and is used for carrying out video monitoring on the vehicle parking process in the parking lot; the second computer image processing module 32 relates to a real-time video target detection algorithm and a key frame comparison technology, receives parking space sideline information detected by a camera, and identifies vehicles which are too close to the sideline and have potential safety hazards; the third output module 33 is configured to output the safety precaution result to the database and management system 4 and the information display device 5, and is configured to input the management system and turn on/off the speaker in the parking lot.
As shown in fig. 3, the video detection and early warning apparatus 3 specifically includes the following steps:
s1: the camera equipment in the parking lot is connected with the gateway equipment, tracks and identifies moving vehicles and personnel, and sends digital signals of parking state information and image information to the computer background system;
s2: the computer receives the monitoring information from the camera equipment and positions and tracks the vehicle which is going to enter or exit the garage by combining the dynamic target recognition algorithm of the computer image processing module 12 of the parking space detection device 1.
S3: and in the real-time video target detection, a Canny edge detection algorithm under a static background is adopted for edge detection.
The Canny edge detection algorithm comprises three steps of filtering processing, edge enhancement and thresholding;
1) and carrying out gray level processing on the input image, and carrying out weighted average according to sampling values of all channels of the image.
2) The filtering process adopts Gaussian filtering, a group of normalized Gaussian kernels are generated by adopting a discretized Gaussian function, and then each point of the image gray matrix is subjected to weighted summation based on the Gaussian kernels, so that the performance of the edge detector is improved.
3) In the edge enhancement process, points with significant changes in the intensity values of the neighborhood of the gray points of the image are highlighted by calculating the gradient amplitude, and the change values of the neighborhood intensity of each point of the image are determined.
4) And discarding points with larger gradient values but not edge points in the neighborhood by adopting a thresholding method.
S4: calculating the deviation position of the vehicle relative to the vehicle line; if the vehicle is too close to the sideline, a loudspeaker of the information display unit is triggered to emit warning sound to warn; if no sideline is too close to the problem, triggering is not carried out;
s5: the video detection early warning result is transmitted to the third output module 33, and the third output module 33 outputs the safety early warning result to the database and management system 4 and the information display device 5 for inputting the management system and switching on and off the loudspeaker in the parking lot.
As shown in fig. 2, the database and management system 4 includes a computer background 41, a first controller module 42, a mobile communication module 43, a motherboard bus 44 and a database system 45; the first controller module 42 and the mobile communication module 43 are in communication connection and information interaction with the motherboard bus 44 through the computer background 41;
the first controller module 42 is used for acquiring, analyzing, processing, coordinating and controlling information of each functional unit in the parking lot vacancy detection and optimal path guiding device; the mobile communication module 43 is used for transmitting parking space detection and parking state information, running state information and running log information of the optimal path guiding device to a remote management center for storage and emergency personnel emergency intervention management; the motherboard bus 44 is used for performing communication connection and information interaction with each functional unit in the parking lot vacancy detection and optimal path guiding device; the database system 45 is used for performing data updating and entry on information output by each functional unit in the parking lot vacancy detection and optimal path guiding device.
As shown in fig. 2, the information display device 5 includes an LED display module 51, a speaker module 52, and a second controller module 53; the LED display module 51 and the loudspeaker module 52 are in communication connection and information interaction with each functional unit in the controller and parking lot vacancy detection and optimal path guiding device through the motherboard bus 44;
the LED display module 51 comprises a parking lot large screen and a ceiling guide rail indicating lamp, wherein the parking lot large screen is used for displaying the current parking space condition in the current parking lot, and the ceiling guide rail indicating lamp is used for guiding the route of the vehicle entering the parking lot according to the calculated optimal path through light color display; the speaker module 52 is used for performing safe voice reminding on the parked vehicle; the second controller module 53 is configured to perform information acquisition, analysis, processing, coordination, and control on an optimal path guiding unit and a video detection early warning unit in the parking lot vacancy detection and optimal path guiding apparatus.
The database and management system 4 and the database system 45 adopt a Microsoft SQL Server database management tool, have the advantages of high-performance design, strong data copying function and management tool and the like, support a Windows graphical management tool, and support local and remote system management and configuration. Meanwhile, the integration degree of related software is high, and the software can be used across various platforms.
The first computer image processing module 12 writes codes through Python to perform image processing.
The entrance node license plate information acquisition module 21 adopts the department service license plate recognition software, and is combined with the entrance node camera equipment, the transmission equipment and the network interface to remotely monitor in real time at any time and is combined with the database and management system 4 and the database information matching module 22 for use.
The database information matching module 22 implements matching and partition guidance of vehicle information by writing a data screening program and a deep learning algorithm through Python.
Claims (3)
1. A parking lot vacancy detection and optimal path guiding device based on image processing comprises a parking lot detection device (1), an optimal path guiding device (2), a video detection early warning device (3), a database and management system (4) and an information display device (5);
the database and management system (4) is used for carrying out coordination management and control on information interaction and transmission among the parking space detection device (1), the optimal path guide device (2), the video detection early warning device (3) and the information display device (5);
the parking space detection device (1) is used for identifying the existing parking condition and the number and positions of empty parking spaces in the parking lot and simultaneously outputting the parking conditions and the number and positions to the database and management system (4) for real-time scheduling;
the optimal path guiding device (2) is used for carrying out region division and optimal path guiding on vehicles entering the parking lot by using a data processing and deep learning algorithm so as to ensure that the optimal path is matched for the entering vehicles and be beneficial to reasonable planning and utilization of large parking lot partitions;
the video detection early warning device (3) is used for carrying out safety early warning on dynamic vehicle targets in the parking lot so as to avoid the obstruction and the scraping of other vehicles caused by improper warehousing directions;
the information display device (5) is used for receiving the path guide information and the early warning information output by the database and management system (4), outputting the path guide information and the early warning information through a large screen of the parking lot, a guide rail indicator lamp of a ceiling and a loudspeaker and reminding driving;
the parking space detection device (1) comprises an image acquisition module (11), a first computer image processing module (12) and a first output module (13); the image acquisition module comprises camera equipment, gateway equipment and transmission equipment connected with the camera equipment and is used for carrying out video monitoring on the parking lot by setting the frame number; the first computer image processing module (12) relates to a real-time video target detection algorithm, dynamic target identification, key frame comparison and camera image receiving for unification processing; the first output module (13) is used for outputting the parking space detection result to the database and management system (4);
the optimal path guiding device (2) comprises an entrance node license plate information acquisition module (21), a database information matching module (22) and a second output module (23); the entrance node license plate information acquisition module (21) comprises entrance camera equipment, gateway equipment and transmission equipment connected with the camera equipment and is used for identifying license plate numbers of vehicles entering the entrance of the parking lot; the database information matching module (22) comprises a data processing module and a deep learning algorithm module, and learns the optimal parking area and path by using vehicle historical data; the second output module (23) is used for outputting the data matching result to the database and management system (4) and the information display device (5) and updating the database and displaying the large screen and ceiling guide rail indicator lamp of the parking lot;
the video detection early warning device (3) comprises a video monitoring module (31), a second computer image processing module (32) and a third output module (33); the video monitoring module (31) comprises camera equipment, gateway equipment and transmission equipment connected with the camera equipment and is used for carrying out video monitoring on the vehicle parking process in the parking lot; the second computer image processing module (32) relates to a real-time video target detection algorithm, receives the parking space sideline information detected by the camera, and identifies vehicles which are too close to the sideline and have potential safety hazards; the third output module (33) is used for outputting the safety early warning result to the database and management system (4) and the information display device (5) and inputting the management system and switching on and off the loudspeaker in the parking lot;
the database and management system (4) comprises a computer background (41), a first controller module (42), a mobile communication module (43), a motherboard bus (44) and a database system (45); the first controller module (42) and the mobile communication module (43) are in communication connection and information interaction with the motherboard bus (44) through the computer background (41); the first controller module (42) is used for acquiring, analyzing, processing, coordinating and controlling information of all functional units in the parking lot vacancy detection and optimal path guiding device; the mobile communication module (43) is used for transmitting parking space detection and parking state information, running state information and running log information of the optimal path guiding device to a remote management center for storage and emergency personnel emergency intervention management; the mother board bus (44) is used for carrying out communication connection and information interaction with each functional unit in the parking lot vacancy detection and optimal path guiding device; the database system (45) is used for updating and inputting data of information output by each functional unit in the parking lot vacancy detection and optimal path guiding device;
the information display device (5) comprises an LED display module (51), a loudspeaker module (52) and a second controller module (53); the LED display module (51) and the loudspeaker module (52) are in communication connection and information interaction with each functional unit in the parking lot vacancy detection and optimal path guiding device through a motherboard bus (44) and a second controller module (53); the LED display module (51) comprises a large parking lot screen and a ceiling guide rail indicator lamp, the large parking lot screen is used for displaying the current parking space condition in the current parking lot, and the ceiling guide rail indicator lamp displays the route of the vehicle entering the parking lot according to the calculated optimal route through the color of the lamp; the loudspeaker module (52) is used for carrying out safe voice reminding on the parked vehicle; the second controller module (53) is used for carrying out information acquisition, analysis, processing, coordination and control on an optimal path guiding unit and a video detection early warning unit in the parking lot vacancy detection and optimal path guiding device.
2. The vacancy detection and optimal path guidance method of the image processing-based parking lot vacancy detection and optimal path guidance device according to claim 1, characterized in that:
the detection method of the parking space detection device (1) is as follows:
s1: the camera equipment in the parking lot is connected with the gateway equipment, tracks and identifies moving vehicles and personnel, and sends digital signals of parking space state information and image information to the computer background system;
s2: the computer dynamic target recognition algorithm is designed based on OpenCV, and is used for detecting a moving target under a dynamic background, wherein the three steps are motion estimation, motion compensation and target detection;
1) and (3) motion estimation: finding a unique matching point of each characteristic point of the reference image in the image by adopting a characteristic point matching algorithm, solving a displacement vector through the change of the position of the point between two frames, and then substituting all motion information into a motion model to solve the global motion parameter of the background; firstly, matching the top point of each unit square with the current characteristic point by taking the current characteristic point as a center and taking 4 as a step to obtain the central position of the optimally matched characteristic point; secondly, changing the stride into 2 and 1 to repeat in sequence on the basis;
2) and motion compensation: affine transformation model parameters are used for replacing and estimating motion vectors of all pixel points, a least square method is adopted for solving known characteristic points, and background movement caused by scanning and shaking of the camera is eliminated by means of superposition of a reverse motion compensation frame image and a reference frame image;
3) detecting a target; performing frame difference on the stable sequence of the background image to obtain a moving target area, and realizing dynamic target detection;
s3: the dynamic target detection transmits the identified empty parking space increase and occupation conditions caused by the vehicle warehouse-out and warehouse-in to the key frame extraction module; the key frame extraction module is used for identifying and comparing a dynamic difference frame and a front static frame and a rear static frame in the video acquisition and dynamic identification processes, further confirming the use condition of an empty parking space, and meanwhile outputting a comparison and verification result to the output module;
s4: the first output module (13) outputs the detection result to the database and management module, and updates the database information;
the optimal path guiding method of the optimal path guiding device (2) is as follows:
s1: the camera equipment at the entrance of the parking lot is connected with the gateway equipment, identifies the license plate number of a vehicle entering the entrance of the parking lot, and sends the digital signal of the license plate information to the computer background system;
s2: the computer receives the license plate information from the camera equipment and inputs the input signal into the database;
s3: the data processing comprises fuzzy information deletion, data fitting processing, data reduction, condition query, time processing and database establishment;
s4: if the license plate information is a registered vehicle, the deep learning module fuses the registration information with historical data of a target vehicle and the existing empty vehicle data of the parking lot, outputs the data to a space-time convolution layer, performs convolutional neural network learning to extract parking characteristics, and outputs an optimal parking space and a parking path according to the empty space condition;
s5: if the license plate information is unregistered but historical driving data exists, the deep learning module firstly calls database data to find historical information of the vehicle for parking in the parking lot, wherein the historical information comprises driving frequency, driving time, parking time and the like; the deep learning module fuses historical data of a target vehicle and the existing empty vehicle data of the parking lot, outputs the data to a space-time convolution layer, performs convolutional neural network learning to extract parking characteristics, and outputs an optimal parking space and a parking path by combining with the empty parking space condition;
s6: if the license plate information is the unregistered vehicle and no historical driving data exists, the deep learning module directly utilizes the existing empty vehicle bit data of the parking lot to output shortest path planning for the driving vehicle so that a driver can find an empty vehicle space as soon as possible;
s7: the parking space allocation result is transmitted to a second output module (23); the second output module (23) outputs the data matching result to the database and management system (4) and the information display device (5) for updating the database and displaying the large screen and ceiling guide rail indicator lamp of the parking lot;
the video detection early warning method of the video detection early warning device (3) is as follows:
s1: the camera equipment in the parking lot is connected with the gateway equipment, tracks and identifies moving vehicles and personnel, and sends digital signals of parking state information and image information to the computer background system;
s2: the computer receives monitoring information from the camera equipment, and a vehicle which is going to enter or exit the garage is positioned and tracked by combining a dynamic target identification algorithm of a computer image processing module (12) of the parking space detection device (1);
s3: in the real-time video target detection, a Canny edge detection algorithm under a static background is adopted for edge detection;
s4: calculating the deviation position of the vehicle relative to the vehicle line; if the vehicle is too close to the sideline, a loudspeaker of the information display unit is triggered to emit warning sound to warn; if no sideline is too close to the problem, triggering is not carried out;
s5: the video detection early warning result is transmitted to a third output module (33), and the third output module (33) outputs the safety early warning result to a database and management system (4) and an information display device (5) for inputting the management system and opening and closing a loudspeaker in the parking lot.
3. The vacancy detection and optimal path guidance method for a parking lot vacancy detection and optimal path guidance apparatus based on image processing according to claim 2, characterized in that: the Canny edge detection algorithm comprises three steps of filtering processing, edge enhancement and thresholding;
1) carrying out gray level processing on an input image, and carrying out weighted average according to sampling values of all channels of the image;
2) the filtering process adopts Gaussian filtering, a group of normalized Gaussian kernels are generated by adopting a discretized Gaussian function, and then each point of the image gray matrix is subjected to weighted summation based on the Gaussian kernels for improving the performance of the edge detector;
3) in the edge enhancement process, points with significant changes in the intensity values of the neighborhood of the gray points of the image are highlighted by calculating gradient amplitude values, and the change values of the neighborhood intensity of each point of the image are determined;
4) and discarding points with larger gradient values but not edge points in the neighborhood by adopting a thresholding method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011235308.6A CN112419778A (en) | 2020-11-09 | 2020-11-09 | Parking lot vacancy detection and optimal path guiding device based on image processing and vacancy detection and optimal path guiding method thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011235308.6A CN112419778A (en) | 2020-11-09 | 2020-11-09 | Parking lot vacancy detection and optimal path guiding device based on image processing and vacancy detection and optimal path guiding method thereof |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112419778A true CN112419778A (en) | 2021-02-26 |
Family
ID=74782260
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011235308.6A Pending CN112419778A (en) | 2020-11-09 | 2020-11-09 | Parking lot vacancy detection and optimal path guiding device based on image processing and vacancy detection and optimal path guiding method thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112419778A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117831336A (en) * | 2023-12-18 | 2024-04-05 | 广东安快智能科技有限公司 | Parking space guiding method, device and equipment based on barrier gate system and storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102117549A (en) * | 2009-12-31 | 2011-07-06 | 北京众库传媒广告有限公司 | Carport guide system used for parking lot |
| CN103810891A (en) * | 2012-11-06 | 2014-05-21 | 西安景行数创信息科技有限公司 | Intelligent garage parking guidance system |
| CN108305504A (en) * | 2018-03-13 | 2018-07-20 | 京东方科技集团股份有限公司 | A kind of parking stall air navigation aid and system |
| CN111311952A (en) * | 2020-02-27 | 2020-06-19 | 平安国际智慧城市科技股份有限公司 | Control method and device for vehicle parking, computer equipment and storage medium |
| CN111627246A (en) * | 2020-06-12 | 2020-09-04 | 桂林电子科技大学 | Intelligent parking space recommendation method based on image recognition and user preference |
-
2020
- 2020-11-09 CN CN202011235308.6A patent/CN112419778A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102117549A (en) * | 2009-12-31 | 2011-07-06 | 北京众库传媒广告有限公司 | Carport guide system used for parking lot |
| CN103810891A (en) * | 2012-11-06 | 2014-05-21 | 西安景行数创信息科技有限公司 | Intelligent garage parking guidance system |
| CN108305504A (en) * | 2018-03-13 | 2018-07-20 | 京东方科技集团股份有限公司 | A kind of parking stall air navigation aid and system |
| CN111311952A (en) * | 2020-02-27 | 2020-06-19 | 平安国际智慧城市科技股份有限公司 | Control method and device for vehicle parking, computer equipment and storage medium |
| CN111627246A (en) * | 2020-06-12 | 2020-09-04 | 桂林电子科技大学 | Intelligent parking space recommendation method based on image recognition and user preference |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117831336A (en) * | 2023-12-18 | 2024-04-05 | 广东安快智能科技有限公司 | Parking space guiding method, device and equipment based on barrier gate system and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109686109B (en) | Parking lot safety monitoring management system and method based on artificial intelligence | |
| KR101942491B1 (en) | Hybrid ai cctv mediation module device consisting of road traffic situation monitoring and real time traffic information analysis | |
| US20220019810A1 (en) | Object Monitoring System and Methods | |
| CN110766975B (en) | Intelligent vehicle searching management method and system for parking lot | |
| KR102122859B1 (en) | Method for tracking multi target in traffic image-monitoring-system | |
| CN105744232A (en) | Method for preventing power transmission line from being externally broken through video based on behaviour analysis technology | |
| KR102174556B1 (en) | Apparatus for monitoring image to control traffic information employing Artificial Intelligence and vehicle number | |
| US20200202142A1 (en) | Autonomous taxi | |
| DE102016215931B4 (en) | System and method for retrievably providing and updating classification data on parking spaces | |
| CN112771858A (en) | Camera assessment techniques for automated vehicles | |
| US11200435B1 (en) | Property video surveillance from a vehicle | |
| US11403947B2 (en) | Systems and methods for identifying available parking spaces using connected vehicles | |
| CN104361764A (en) | Stereo garage managing device and method based on video monitoring | |
| CN111932934A (en) | Intelligent parking lot in-and-out vehicle management system based on big data | |
| CN110363983B (en) | Non-signalized intersection unmanned and non-manned real-time trajectory prediction and conflict recognition method | |
| CN110175533A (en) | Overpass traffic condition method of real-time, device, terminal and storage medium | |
| CN110473428A (en) | A kind of intelligent parking method, apparatus and system | |
| CN110263623A (en) | Train climbs monitoring method, device, terminal and storage medium | |
| CN114022767A (en) | Elevator floor positioning method and device, terminal equipment and storage medium | |
| CN114677848B (en) | Perception early warning system, method, device and computer program product | |
| CN112419778A (en) | Parking lot vacancy detection and optimal path guiding device based on image processing and vacancy detection and optimal path guiding method thereof | |
| CN117292322A (en) | Deep learning-based personnel flow detection method and system | |
| CN111469765A (en) | Detection system with function of removing vehicle body shadows and method thereof | |
| CN111739332A (en) | Parking lot management system | |
| CN114549444A (en) | Cross-camera target detection and tracking system and method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210226 |