CN109410635B

CN109410635B - Roadside parking space state identification method and device

Info

Publication number: CN109410635B
Application number: CN201811260332.8A
Authority: CN
Inventors: 王学彬; 班华忠; 尹方始; 李党
Original assignee: Beijing Icetech Science & Technology Co ltd
Current assignee: Beijing Icetech Science & Technology Co ltd
Priority date: 2018-10-26
Filing date: 2018-10-26
Publication date: 2021-01-05
Anticipated expiration: 2038-10-26
Also published as: CN109410635A

Abstract

The invention provides a roadside parking space state identification method, which comprises the following steps: inputting or collecting roadside video images; acquiring a shielding parameter, a parking space area, an interested area and a non-parking space area, identifying a license plate, and taking the identified license plate as a license plate to be matched; matching the license plate to be matched with the current parking space license plate, allocating the matched license plate to a corresponding parking space, respectively matching the current non-parking space license plate with the parking space license plate, and the residual unassigned license plate with the non-parking space license plate, and allocating the license plate according to the matching result; when the shielding parameter is 0, performing first or second counting processing on the current parking space; when the shielding parameter is 1, performing first or second counting processing on the current parking space; updating the current parking space state; and outputting the parking space states and the license plate information of all the parking spaces. Compared with the prior art, the method can effectively solve the problem of license plate shielding in the roadside parking space state identification, and identify the parking space state.

Description

Roadside parking space state identification method and device

Technical Field

The invention relates to image processing, video monitoring and intelligent parking, in particular to a method and a device for recognizing the state of a roadside parking space.

Background

With the increasing amount of automobiles, roadside parking also becomes one of the problems to be solved urgently, and the importance of the traffic management department has been brought forward. Traditional roadside parking detection is mainly that enclosure, ground feel device and handheld machine, and is inefficient, can't realize real time monitoring, has greatly wasted manpower and financial resources of relevant departments.

In recent years, a method for roadside parking detection based on a monitoring video receives great attention, and has the advantages of high accuracy, good real-time performance, low cost, easiness in evidence collection and the like. According to the installation position of the monitoring camera, a roadside low pole scheme (the installation height is 0.8-1.8 m) and a roadside high pole scheme (the installation height is 3.5-7 m) are provided. The low pole scheme is that a camera is arranged on a low pole to monitor the state of a parking space; the high-pole scheme is that a plurality of guns and a plurality of ball machines are combined, and the guns and the ball machines are linked to realize the detection of a plurality of (2-12) parking spaces.

However, the problem of license plate shielding cannot be effectively solved by the existing parking space state recognition scheme.

In summary, there is an urgent need to provide a method and a device for recognizing the state of a roadside parking space, which can effectively solve the problem of license plate shielding.

Disclosure of Invention

In view of this, the main object of the present invention is to provide a roadside parking space state identification scheme, which can effectively solve the problem of license plate occlusion.

In order to achieve the above object, according to a first aspect of the present invention, there is provided a roadside parking space state identification method, including:

the method comprises the following steps of firstly, inputting or collecting a roadside video image;

secondly, acquiring shielding parameters, parking space areas, interested areas and non-parking space areas, identifying license plates of the interested areas, and taking the identified license plates as license plates to be matched;

matching license plates to be matched with all the parking place license plates, distributing the matched license plates to corresponding parking places, acquiring unallocated license plates, respectively matching the current non-parking place license plate with the parking place license plate, and the residual unallocated license plates with the non-parking place license plates, and distributing the license plates according to matching results;

a fourth step of performing first counting processing on the current parking space if the license plate is allocated to the current parking space and the allocated license plate is not in the current parking space area when the shielding parameter is 0, and performing second counting processing otherwise;

fifthly, when the shielding parameter is 1, performing first counting processing or second counting processing on the current parking space according to whether the current parking space is allocated with a license plate, whether the allocated license plate is in the current parking space area and is matched with the license plate of the previous parking space;

a sixth step of updating the current parking space state according to the current parking space state, the vehicle-presence counting value, the vehicle-absence counting value, the vehicle-presence time and the vehicle-absence time;

and a seventh step of outputting the parking space states and license plate information of all parking spaces.

Further, the second step includes:

a shielding parameter obtaining step, wherein if the parking space type is a vertical parking space, the shielding parameter is set to be 1, otherwise, the shielding parameter is set to be 0;

a parking space region acquisition step, namely dividing a quadrilateral region according to a parking space line of the roadside video image, taking the quadrilateral region as a parking space region, and numbering the parking space regions from bottom to top;

a step of obtaining the region of interest, namely obtaining the circumscribed rectangular regions of all parking space regions, and respectively extending the upper and lower boundaries of the circumscribed rectangular regions upwards and downwards to the height k of the circumscribed rectangular regions₁、k₂Multiplying, expanding the left and right boundaries of the circumscribed rectangular region to the left and right by k of the width of the circumscribed rectangular region₃、k₄Multiplying, wherein the obtained expansion area is used as an interested area;

a non-parking space region acquisition step, wherein a region which does not belong to a parking space region in the region of interest is used as a non-parking space region;

a license plate recognition step, wherein license plate recognition is carried out on the region of interest to obtain a recognized license plate and a position;

and a license plate to be matched outputting step, namely taking the identified license plate as a license plate to be matched, and outputting the license plate to be matched and the corresponding position.

Further, the region of interest acquiring step includes:

an external rectangular area obtaining step, namely respectively counting boundary values of upper, lower, left and right boundaries of all parking space areas, taking the minimum boundary value of the left boundary and the minimum boundary value of the upper boundary as the left boundary and the upper boundary of the external rectangular area, and taking the maximum boundary value of the right boundary and the lower boundary as the right boundary and the lower boundary of the external rectangular area;

expanding the circumscribed rectangular region according to the upper boundary y of the circumscribed rectangular region₁And the lower boundary y is y₂X is left boundary₁And the right boundary x ═ x₂Acquiring an upper boundary y ═ y of the region of interest₁-k₁h. Lower boundary y ═ y₂+k₂h. Left border x ═ x₁-k₃w, right boundary x ═ x₂+k₄w, wherein k₁、k₂、k₃、k₄And h and w are the height and width of the circumscribed rectangular region, and if the boundary of the region of interest exceeds the boundary of the roadside video image, the exceeded boundary is set as the boundary of the corresponding roadside video image.

Further, the third step includes:

the method comprises the following steps of a first matching step, matching license plates to be matched with license plates of all parking spaces, allocating the matched license plates to corresponding parking spaces, and marking the unmatched license plates as the unmatched license plates; judging whether the position of the license plate which is not distributed is in a certain parking space area or not, and distributing the license plate which is not distributed to the parking space of the parking space area;

a second matching step, matching the current non-parking space license plate with the parking space license plate, endowing the matched non-parking space license plate with a corresponding parking space, and deleting the information of the non-parking space license plate;

and a third matching step, matching the residual unallocated license plates with the non-parking space license plates, allocating the matched unallocated license plates to the corresponding non-parking spaces, and establishing new non-parking space license plate information for the unmatched unallocated license plates.

Further, the fifth step includes:

and (3) processing the allocated license plate, if the license plate is allocated to the current parking space, performing the following processing: if the license plate is not in the parking space area, performing first counting processing on the current parking space; if the license plate is in the parking space area and the previous parking space is allocated to the license plate, performing second counting processing on the current parking space; if the license plate is in the parking space area, the previous parking space is not allocated with the license plate, but historical license plate information exists, and the historical license plate information is matched with the current parking space, performing first counting processing on the previous parking space, and performing second counting processing on the current parking space;

and (3) processing the unallocated license plate, if the license plate is not allocated to the current parking space, performing the following processing: and if the current parking space does not have the historical license plate information, processing according to the step of no historical license plate processing, and otherwise, processing according to the step of historical license plate processing.

Further, the no history license plate processing step comprises: and if the no-vehicle counting value of the current parking space is greater than 0, performing first counting processing on the current parking space, otherwise, performing second counting processing on the current parking space.

The processing steps of the license plate with history comprise: if the previous parking space does not have license plate information, if the current parking space has no license plate information, the current parking space is subjected to first counting processing, and if not, the current parking space is subjected to second counting processing; if the previous parking space has license plate information and the current parking space is matched with the license plate of the previous parking space, if the previous parking space is allocated to the license plate or the number of the current parking space is greater than 0, performing first counting processing on the current parking space, otherwise, performing second counting processing on the current parking space; the previous parking space has license plate information, the license plate of the current parking space is not matched with that of the previous parking space, if the number of the non-vehicle counting value is larger than 0, the current parking space is subjected to first counting processing, and if not, the current parking space is subjected to second counting processing.

Further, the first counting process is: counting the number of the vehicles is clear 0, and counting the number of the vehicles without the vehicles is added with 1; the second counting process is as follows: the no-vehicle counting value is cleared to be 0, and the vehicle counting value is added with 1. The initial value of the no-vehicle counting value and the vehicle counting value of the parking space is 0.

Further, the sixth step includes:

a parking space unknown state updating step, wherein when the current parking space state is unknown, if the number of the current parking space with cars is larger than a first threshold value, the current parking space state is updated to be driven-in, if the number of the current parking space without cars is larger than the first threshold value, the current parking space state is updated to be driven-out, otherwise, the current parking space state is not updated;

a parking space driving state updating step, counting the time of a license plate in the current parking space when the current parking space state is driving in, updating the current parking space state to be occupied if the time of the license plate is larger than a second threshold value, updating the current parking space state to be driven out if the no-vehicle counting value of the current parking space is larger than a first threshold value, and not updating the current parking space state if the no-vehicle counting value of the current parking space is larger than the first threshold value;

a parking space occupation state updating step, wherein when the current parking space state is occupied, if the no-vehicle counting value of the current parking space is larger than a first threshold value, the current parking space state is updated to an outgoing state, otherwise, the current parking space state is not updated;

a parking space exit state updating step, counting the time when no license plate exists in the current parking space when the current parking space state is exit, updating the current parking space state to be idle if the time when no license plate exists is greater than a second threshold, updating the current parking space state to be an entry state if the number of vehicles in the current parking space is greater than a first threshold, and otherwise not updating the current parking space state;

and a parking space idle state updating step, wherein when the current parking space state is idle, if the number of the current parking space is greater than a first threshold value, the current parking space state is updated to be driven in, otherwise, the current parking space state is not updated.

According to another aspect of the present invention, there is provided a roadside parking space state recognition device, including:

the video image input or acquisition module is used for inputting or acquiring a roadside video image;

the license plate acquisition module is used for acquiring the shielding parameters, the parking space area, the interested area and the non-parking space area, identifying the license plate of the interested area and taking the identified license plate as the license plate to be matched;

the license plate distribution module is used for matching license plates to be matched with license plates of all parking spaces, distributing the matched license plates to corresponding parking spaces, acquiring unallocated license plates, respectively matching the current non-parking space license plate with the parking space license plate, and the residual unallocated license plates with the non-parking space license plates, and distributing the license plates according to the matching result;

the 0 shielding parameter processing module is used for performing first counting processing on the current parking space if the current parking space is allocated with a license plate and the allocated license plate is not in the current parking space area when the shielding parameter is 0, and performing second counting processing if the current parking space is not allocated with the license plate;

the system comprises a 1 shielding parameter processing module, a first counting processing module and a second counting processing module, wherein the 1 shielding parameter processing module is used for carrying out first counting processing or second counting processing on a current parking space according to whether a license plate is allocated to the current parking space, whether the allocated license plate is in a current parking space area and whether the allocated license plate is matched with a license plate of a previous parking space;

the parking space state updating module is used for updating the current parking space state according to the current parking space state, the vehicle-presence counting value, the vehicle-absence counting value, the vehicle license plate-presence time and the vehicle license plate-absence time;

and the parking space state and license plate information output module is used for outputting the parking space states and license plate information of all parking spaces.

Further, the region of interest and license plate to be matched acquisition module comprises:

the shielding parameter acquisition module is used for setting the shielding parameter to be 1 if the parking space type is a vertical parking space, or setting the shielding parameter to be 0;

the parking space region acquisition module is used for dividing a quadrilateral region according to a parking space line of the roadside video image, taking the quadrilateral region as a parking space region and numbering the parking space regions from bottom to top;

the interesting region acquisition module is used for acquiring the circumscribed rectangular regions of all the parking space regions and respectively extending the upper and lower boundaries of the circumscribed rectangular regions upwards and downwards to the height k of the circumscribed rectangular regions₁、k₂Multiplying, expanding the left and right boundaries of the circumscribed rectangular region to the left and right by k of the width of the circumscribed rectangular region₃、k₄Multiplying, wherein the obtained expansion area is used as an interested area;

the non-parking space region acquisition module is used for taking a region which does not belong to a parking space region in the region of interest as a non-parking space region;

the license plate recognition module is used for recognizing the license plate of the region of interest to acquire the recognized license plate and position;

and the license plate output module to be matched is used for outputting the license plate to be matched and the corresponding position by taking the identified license plate as the license plate to be matched.

Further, the license plate assignment module includes:

the first matching module is used for matching the license plates to be matched with the license plates of all the parking spaces, allocating the matched license plates to the corresponding parking spaces and marking the unmatched license plates as the unmatched license plates; judging whether the position of the license plate which is not distributed is in a certain parking space area or not, and distributing the license plate which is not distributed to the parking space of the parking space area;

the second matching module is used for matching the current non-parking space license plate with the parking space license plate, endowing the matched non-parking space license plate with a corresponding parking space, and deleting the information of the non-parking space license plate;

and the third matching module is used for matching the residual unallocated license plates with the non-parking space license plates, allocating the matched unallocated license plates to the corresponding non-parking spaces, and establishing new non-parking space license plate information for the unmatched unallocated license plates.

Further, the 1 occlusion parameter processing module includes:

the allocated license plate processing module is used for performing the following processing if the license plate is allocated to the current parking space: if the license plate is not in the parking space area, performing first counting processing on the current parking space; if the license plate is in the parking space area and the previous parking space is allocated to the license plate, performing second counting processing on the current parking space; if the license plate is in the parking space area, the previous parking space is not allocated with the license plate, but historical license plate information exists, and the historical license plate information is matched with the current parking space, performing first counting processing on the previous parking space, and performing second counting processing on the current parking space;

and the unallocated license plate processing module is used for performing the following processing if the license plate is not allocated to the current parking space: and if the current parking space does not have the historical license plate information, processing according to the historical license plate processing module, and otherwise, processing according to the historical license plate processing module.

Wherein, no historical license plate processing module includes: and if the no-vehicle counting value of the current parking space is greater than 0, performing first counting processing on the current parking space, and otherwise, performing second counting processing on the current parking space.

The historical license plate processing module comprises: the system is used for performing first counting processing on the current parking space if the previous parking space has no license plate information, and if the current parking space has no license plate information, the counting value is greater than 0, and otherwise, performing second counting processing on the current parking space; if the previous parking space has license plate information and the current parking space is matched with the license plate of the previous parking space, if the previous parking space is allocated to the license plate or the number of the current parking space is greater than 0, performing first counting processing on the current parking space, otherwise, performing second counting processing on the current parking space; the previous parking space has license plate information, the license plate of the current parking space is not matched with that of the previous parking space, if the number of the non-vehicle counting value is larger than 0, the current parking space is subjected to first counting processing, and if not, the current parking space is subjected to second counting processing.

Further, the parking space state updating module comprises:

the parking place unknown state updating module is used for updating the current parking place state into driving-in if the number of the current parking place with the vehicle is larger than a first threshold value when the current parking place state is unknown, updating the current parking place state into driving-out if the number of the current parking place without the vehicle is larger than the first threshold value, and otherwise, not updating the current parking place state;

the parking space driving-in state updating module is used for counting the time when a license plate exists in the current parking space when the current parking space state is driving-in, updating the current parking space state to be occupied if the time when the license plate exists is larger than a second threshold value, updating the current parking space state to be driven-out if the no-vehicle counting value of the current parking space is larger than a first threshold value, and otherwise, not updating the current parking space state;

the parking space occupation state updating module is used for updating the current parking space state into an outgoing state if the no-vehicle counting value of the current parking space is greater than a first threshold value when the current parking space state is occupied, and otherwise, the current parking space state is not updated;

the parking space exit state updating module is used for counting the time when the current parking space state is exit, updating the current parking space state to be idle if the time when no license plate exists is greater than a second threshold value, updating the current parking space state to be an entry state if the number of vehicles in the current parking space is greater than a first threshold value, and otherwise not updating the current parking space state;

and the parking space idle state updating module is used for updating the current parking space state into a driving-in state if the current parking space state is idle and the vehicle number of the current parking space is greater than a first threshold value, otherwise, the current parking space state is not updated.

Compared with the existing roadside parking space state identification technology, the roadside parking space state identification method and device respectively perform parking space license plate allocation processing according to the shielding parameter 0 and the shielding parameter 1, meanwhile, the parking space state is updated in real time, the problem of license plate shielding can be effectively solved, and 5 states of parking spaces can be identified in real time.

Drawings

Fig. 1 shows a flowchart of a roadside parking space state identification method according to the present invention.

Fig. 2 shows a block diagram of a roadside parking space state recognition device according to the invention.

Detailed Description

To further clarify the structure, characteristics and other objects of the present invention, those skilled in the art will now describe in detail the preferred embodiments of the present invention with reference to the attached drawings, which are provided for the purpose of describing the technical solutions of the present invention only and are not intended to limit the present invention.

Fig. 1 shows a flowchart of a roadside parking space state identification method according to the present invention. As shown in fig. 1, a roadside parking space state identification method according to the present invention includes:

a first step S1 of inputting or acquiring a roadside video image;

a second step S2 of obtaining the shielding parameters, the parking space area, the interested area and the non-parking space area, carrying out license plate recognition on the interested area, and taking the recognized license plate as a license plate to be matched;

the third step S3, matching the license plates to be matched with all the parking spaces, allocating the matched license plates to corresponding parking spaces, acquiring unallocated license plates, respectively matching the current non-parking space license plates with the parking space license plates and the residual unallocated license plates with the non-parking space license plates, and allocating the license plates according to the matching result;

a fourth step S4, when the shielding parameter is 0, if the license plate is allocated to the current parking space and the allocated license plate is not in the current parking space area, performing a first counting process on the current parking space, otherwise, performing a second counting process;

a fifth step S5, when the shielding parameter is 1, performing a first counting process or a second counting process on the current parking space according to whether the license plate is allocated to the current parking space, whether the allocated license plate is in the current parking space area, and whether the license plate is matched with the license plate of the previous parking space;

a sixth step S6 of updating the current parking space state according to the current parking space state, the vehicle-presence count value, the vehicle-absence count value, the vehicle license plate-presence time and the vehicle license plate-absence time;

and a seventh step S7 of outputting the parking space states and license plate information of all the parking spaces.

The roadside video image in the first step S1 is a video image covering a roadside parking space region range, and may be directly input or acquired through image acquisition equipment.

Further, image acquisition equipment installs in the roadside control pole setting, and the mounting height is located 5 ~ 8 meters on ground, can cover 2 ~ 8 parking stall ranges. Exemplarily, the first step S1 is: through installing the image acquisition equipment in the roadside monitoring pole setting, gather roadside video image, wherein mounting height is located 6 meters on ground, can cover 4 parking stall scopes.

Further, the second step S2 includes:

an occlusion parameter obtaining step S21, if the parking space type is a vertical parking space, setting the occlusion parameter as 1, otherwise, setting the occlusion parameter as 0;

a parking space region obtaining step S22 of dividing a quadrilateral region according to a parking space line of the roadside video image, taking the quadrilateral region as a parking space region, and numbering the parking space regions in a sequence from bottom to top;

an interested area obtaining step S23, obtaining the circumscribed rectangular areas of all parking space areas, and respectively extending the upper and lower boundaries of the circumscribed rectangular areas upwards and downwards to the height k of the circumscribed rectangular areas₁、k₂Multiplying, expanding the left and right boundaries of the circumscribed rectangular region to the left and right by k of the width of the circumscribed rectangular region₃、k₄Multiplying, wherein the obtained expansion area is used as an interested area;

a non-parking space region acquisition step S24, wherein a region which does not belong to a parking space region in the region of interest is used as a non-parking space region;

a license plate recognition step S25, recognizing the license plate of the region of interest to obtain a recognized license plate and a position;

and a license plate to be matched output step S26, wherein the recognized license plate is used as a license plate to be matched, and the license plate to be matched and the corresponding position are output.

The shielding parameter obtaining step S21 may perform manual configuration according to the parking space type in the scene, where if the parking space type in the scene is a vertical parking space, the shielding parameter is set to 1, and otherwise, the shielding parameter is set to 0.

In the parking space region obtaining step S22, the manual quadrilateral area may be drawn as the parking space region according to the parking space line of the roadside video image, or the parking space line of the roadside video image may be obtained according to the parking space line detection method, and the quadrilateral area formed by the parking space lines is used as the parking space region; and numbering the parking space regions according to the sequence of the roadside video images from bottom to top. Illustratively, according to the line of the roadside video image, 4 quadrilateral areas are artificially drawn as parking space areas, and are numbered as the 1 st parking space area, the 2 nd parking space area, the 3 rd parking space area and the 4 th parking space area from bottom to top in sequence.

Further, the region of interest acquiring step S23 includes:

an external rectangular region obtaining step S231, which respectively counts boundary values of upper, lower, left and right boundaries of all parking space regions, takes the minimum boundary value of the left boundary and the minimum boundary value of the upper boundary as the left boundary and the upper boundary of the external rectangular region, and takes the maximum boundary value of the right boundary and the lower boundary as the right boundary and the lower boundary of the external rectangular region;

step S232 of expanding the circumscribed rectangular region according to the upper boundary y of the circumscribed rectangular region₁And the lower boundary y is y₂X is left boundary₁And the right boundary x ═ x₂Acquiring an upper boundary y ═ y of the region of interest₁-k₁h. Lower boundary y ═ y₂+k₂h. Left border x ═ x₁-k₃w, right boundary x ═ x₂+k₄w, wherein k₁、k₂、k₃、k₄And h and w are the height and width of the circumscribed rectangular region, and if the boundary of the region of interest exceeds the boundary of the roadside video image, the exceeded boundary is set as the boundary of the corresponding roadside video image.

Illustratively, the circumscribed rectangular area obtaining step S231 is: if there are 4 parking space areasThe minimum boundary value x of the left boundary of the 4 parking space areas is counted_1minMaximum boundary value x of right boundary_2maxMinimum boundary value y of the upper boundary_1minMaximum boundary value y of lower boundary_2maxLet x be left boundary ═ x_1minX is the right boundary_2maxThe upper boundary is y ═ y_1minThe lower boundary is y ═ y_2maxThe formed area is used as a circumscribed rectangular area.

Further, k is₁、k₂、k₃、k₄The value range of (A) is 0.1-0.6.

Illustratively, the step S232 of expanding the circumscribed rectangular region is: according to the upper boundary y of the circumscribed rectangular region₁And the lower boundary y is y₂X is left boundary₁And the right boundary x ═ x₂Acquiring an upper boundary y ═ y of the region of interest₁-0.4 xh, lower boundary y ═ y₂+0.4 xh, left border x ═ x₁-0.3 xw, right border x ═ x₂+0.3 xw. When the upper boundary of the region of interest is smaller than 0, setting the upper boundary of the region of interest as the upper boundary of the roadside video image; when the lower boundary of the region of interest is larger than the height of the roadside video image, setting the lower boundary of the region of interest as the lower boundary of the roadside video image; when the left boundary of the region of interest is smaller than 0, setting the left boundary of the region of interest as the left boundary of the roadside video image; and when the right boundary of the region of interest is larger than the width of the roadside video image, setting the right boundary of the region of interest as the right boundary of the roadside video image.

The license plate recognition step S25 may be implemented by using an existing license plate recognition method, and includes one or a combination of the following methods: the system comprises a license plate recognition method based on a classifier, a license plate recognition method based on a neural network and a license plate recognition method based on deep learning.

Further, the third step S3 includes:

a first matching step S31, matching license plates to be matched with license plates of all parking spaces, allocating the matched license plates to corresponding parking spaces, and marking the unmatched license plates as the unmatched license plates; judging whether the position of the license plate which is not distributed is in a certain parking space area or not, and distributing the license plate which is not distributed to the parking space of the parking space area;

a second matching step S32, matching the current non-parking space license plate with the parking space license plate, assigning the matched non-parking space license plate to a corresponding parking space, and deleting the information of the non-parking space license plate;

and a third matching step S33, matching the residual unallocated license plates with the non-parking space license plates, allocating the matched unallocated license plates to the corresponding non-parking spaces, and establishing new non-parking space license plate information for the unmatched unallocated license plates.

And the initial values of the parking space license plate and the non-parking space license plate are null.

Further, the license plate matching is to compare license plate numbers one by one, if the license plate numbers are all the same, the license plate numbers are considered to be matched, otherwise, the license plate numbers are considered to be not matched. Illustratively, if the license plate number to be matched is 'lu A8M 123', and the current parking space license plate number is 'lu A8M 123', the license plate to be matched is considered to be matched with the current parking space license plate.

Further, the fifth step S5 includes:

a license plate processing step S51, if the license plate is allocated to the current parking space, the following processing is carried out: if the license plate is not in the parking space area, performing first counting processing on the current parking space; if the license plate is in the parking space area and the previous parking space is allocated to the license plate, performing second counting processing on the current parking space; if the license plate is in the parking space area, the previous parking space is not allocated with the license plate, but historical license plate information exists, and the historical license plate information is matched with the current parking space, performing first counting processing on the previous parking space, and performing second counting processing on the current parking space;

and a license plate unallocated processing step S52, wherein if a license plate is not allocated to the current parking space, the following processing is performed: if the current parking space has no historical license plate information, processing according to the step S521 of no historical license plate processing, and otherwise, processing according to the step S522 of historical license plate processing.

Wherein the no history license plate processing step S521 includes: and if the no-vehicle counting value of the current parking space is greater than 0, performing first counting processing on the current parking space, otherwise, performing second counting processing on the current parking space.

The step S522 of processing the license plate with history includes: if the previous parking space does not have license plate information, if the current parking space has no license plate information, the current parking space is subjected to first counting processing, and if not, the current parking space is subjected to second counting processing; if the previous parking space has license plate information and the current parking space is matched with the license plate of the previous parking space, if the previous parking space is allocated to the license plate or the number of the current parking space is greater than 0, performing first counting processing on the current parking space, otherwise, performing second counting processing on the current parking space; the previous parking space has license plate information, the license plate of the current parking space is not matched with that of the previous parking space, if the number of the non-vehicle counting value is larger than 0, the current parking space is subjected to first counting processing, and if not, the current parking space is subjected to second counting processing.

And the previous parking space is the parking space with the number which is the previous number of the current parking space. For example, if the current parking space is the 4 th parking space, the previous parking space of the current parking space is not the 3 rd parking space.

The historical license plate information of the parking space is license plate information distributed before the current frame of the parking space. Illustratively, for each frame of roadside video images, if a parking space is allocated to a passing license plate, the parking space is stored in a license plate queue of a corresponding parking space to serve as historical license plate information of the corresponding parking space, and if the parking space is not allocated to the passing license plate all the time, the parking space is considered to have no historical license plate information.

Further, the sixth step S6 includes:

a parking space unknown state updating step S61, when the current parking space state is unknown, if the vehicle-containing count value of the current parking space is greater than the first threshold value, updating the current parking space state as drive-in, if the vehicle-free count value of the current parking space is greater than the first threshold value, updating the current parking space state as drive-out, otherwise, not updating the current parking space state;

a parking space driving state updating step S62, when the current parking space state is driving in, counting the time of a license plate in the current parking space, if the time of the license plate is larger than a second threshold value, updating the current parking space state to be occupied, if the no-vehicle count value of the current parking space is larger than a first threshold value, updating the current parking space state to be driving out, otherwise, not updating the current parking space state;

a parking space occupation state updating step S63, when the current parking space state is occupied, if the no-vehicle count value of the current parking space is larger than a first threshold value, updating the current parking space state to a driving-out state, otherwise, not updating the current parking space state;

a parking space exit state updating step S64, when the current parking space state is exit, counting the time of no license plate of the current parking space, if the time of no license plate is greater than a second threshold value, updating the current parking space state to be idle, if the number of vehicles in the current parking space is greater than a first threshold value, updating the current parking space state to be an entry state, otherwise, not updating the current parking space state;

and a parking space idle state updating step S65, when the current parking space state is idle, if the number of vehicles in the current parking space is greater than the first threshold value, updating the current parking space state as drive-in, otherwise, not updating the current parking space state.

Wherein, the initial value of current parking stall state is unknown.

Further, the value range of the first threshold is 1-120 seconds, and the value range of the second threshold is 5-40 seconds. Illustratively, the first threshold is selected to be 10 and the second threshold is selected to be 12 seconds.

The seventh step S7 includes: and outputting the parking space state and the corresponding license plate information of each parking space region in the current frame roadside video image.

Fig. 2 is a block diagram of a roadside parking space state recognition device according to the present invention. As shown in fig. 2, a roadside parking space state recognition device according to the present invention includes:

the video image input or acquisition module 1 is used for inputting or acquiring roadside video images;

the license plate acquisition module 2 is used for acquiring the shielding parameters, the parking space area, the interested area and the non-parking space area, identifying the license plate of the interested area and taking the identified license plate as the license plate to be matched;

the license plate distribution module 3 is used for matching license plates to be matched with license plates of all parking spaces, distributing the matched license plates to corresponding parking spaces, acquiring unallocated license plates, respectively matching the current non-parking space license plate with the parking space license plate, and the residual unallocated license plates with the non-parking space license plates, and distributing the license plates according to the matching result;

the 0 shielding parameter processing module 4 is used for performing first counting processing on the current parking space if the license plate is allocated to the current parking space and the allocated license plate is not in the current parking space area when the shielding parameter is 0, and performing second counting processing otherwise;

the system comprises a 1 shielding parameter processing module 5, a first counting module and a second counting module, wherein the 1 shielding parameter processing module is used for carrying out first counting processing or second counting processing on a current parking space according to whether a license plate is allocated to the current parking space, whether the allocated license plate is in a current parking space area and whether the allocated license plate is matched with a license plate of a previous parking space;

the parking space state updating module 6 is used for updating the current parking space state according to the current parking space state, the vehicle-presence counting value, the vehicle-absence counting value, the vehicle license plate-presence time and the vehicle license plate-absence time;

and the parking space state and license plate information output module 7 is used for outputting the parking space states and license plate information of all parking spaces.

The video image input or acquisition module 1 may be implemented by a video image input device or an image acquisition device.

Further, image acquisition equipment installs in the roadside control pole setting, and the mounting height is located 5 ~ 8 meters on ground, can cover 2 ~ 8 parking stall ranges.

Further, the region of interest and license plate to be matched obtaining module 2 includes:

the shielding parameter acquiring module 21 is configured to set the shielding parameter to 1 if the parking space type is a vertical parking space, and set the shielding parameter to 0 if the parking space type is not a vertical parking space;

the parking space region acquisition module 22 is configured to divide a quadrilateral region according to a parking space line of the roadside video image, use the quadrilateral region as a parking space region, and number the parking space regions in a sequence from bottom to top;

a region-of-interest obtaining module 23, configured to obtain a circumscribed rectangular region of all parking space regions, and extend the upper and lower boundaries of the circumscribed rectangular region upward and downward by k, which is the height of the circumscribed rectangular region, respectively₁、k₂Multiplying, expanding the left and right boundaries of the circumscribed rectangular region to the left and right by k of the width of the circumscribed rectangular region₃、k₄Multiplying, wherein the obtained expansion area is used as an interested area;

a non-parking space region acquisition module 24, configured to use a region that does not belong to a parking space region in the region of interest as a non-parking space region;

the license plate recognition module 25 is used for recognizing the license plate of the region of interest to acquire the recognized license plate and position;

and the license plate output module 26 to be matched is used for outputting the license plate to be matched and the corresponding position by taking the identified license plate as the license plate to be matched.

The shielding parameter obtaining module 21 is configured to perform manual configuration according to the parking space type in the scene, and if the parking space type in the scene is a vertical parking space, the shielding parameter is set to 1, otherwise, the shielding parameter is set to 0.

Further, the region of interest acquisition module 23 includes:

the external rectangular area obtaining module 231 is configured to count boundary values of upper, lower, left and right boundaries of all parking space areas, respectively, use the minimum boundary value of the left boundary and the minimum boundary value of the upper boundary as the left boundary and the upper boundary of the external rectangular area, and use the maximum boundary value of the right boundary and the lower boundary as the right boundary and the lower boundary of the external rectangular area;

a circumscribed rectangular region outward-expanding module 232, configured to expand the upper boundary y of the circumscribed rectangular region to y₁And the lower boundary y is y₂X is left boundary₁And the right boundary x ═ x₂Acquiring an upper boundary y ═ y of the region of interest₁-k₁h. Lower boundary y ═ y₂+k₂h. Left border x ═ x₁-k₃w, right boundary x ═ x₂+k₄w, wherein k₁、k₂、k₃、k₄And h and w are the height and width of the circumscribed rectangular region, and if the boundary of the region of interest exceeds the boundary of the roadside video image, the exceeded boundary is set as the boundary of the corresponding roadside video image.

Further, k is₁、k₂、k₃、k₄The value range of (A) is 0.1-0.6.

Further, the license plate assignment module 3 includes:

the first matching module 31 is used for matching the license plates to be matched with the license plates of all the parking spaces, allocating the matched license plates to the corresponding parking spaces, and marking the unmatched license plates as the unmatched license plates; judging whether the position of the license plate which is not distributed is in a certain parking space area or not, and distributing the license plate which is not distributed to the parking space of the parking space area;

the second matching module 32 is used for matching the current non-parking space license plate with the parking space license plate, endowing the matched non-parking space license plate with a corresponding parking space, and deleting the information of the non-parking space license plate;

and the third matching module 33 is configured to match the remaining unassigned license plates with the non-parking license plates, assign the matched unassigned license plates to corresponding non-parking spaces, and establish new non-parking space license plate information for the unassigned license plates.

Further, the 1-occlusion parameter processing module 5 includes:

the assigned license plate processing module 51 is configured to, if a license plate is assigned to the current parking space, perform the following processing: if the license plate is not in the parking space area, performing first counting processing on the current parking space; if the license plate is in the parking space area and the previous parking space is allocated to the license plate, performing second counting processing on the current parking space; if the license plate is in the parking space area, the previous parking space is not allocated with the license plate, but historical license plate information exists, and the historical license plate information is matched with the current parking space, performing first counting processing on the previous parking space, and performing second counting processing on the current parking space;

the unassigned license plate processing module 52 is configured to, if a license plate is not assigned to the current parking space, perform the following processing: if the current parking space has no historical license plate information, processing is carried out according to the historical license plate processing module 521, and otherwise, processing is carried out according to the historical license plate processing module 522.

Wherein the no history license plate processing module 521 includes: and if the no-vehicle counting value of the current parking space is greater than 0, performing first counting processing on the current parking space, and otherwise, performing second counting processing on the current parking space.

The historical license plate processing module 522 includes: the system is used for performing first counting processing on the current parking space if the previous parking space has no license plate information, and if the current parking space has no license plate information, the counting value is greater than 0, and otherwise, performing second counting processing on the current parking space; if the previous parking space has license plate information and the current parking space is matched with the license plate of the previous parking space, if the previous parking space is allocated to the license plate or the number of the current parking space is greater than 0, performing first counting processing on the current parking space, otherwise, performing second counting processing on the current parking space; the previous parking space has license plate information, the license plate of the current parking space is not matched with that of the previous parking space, if the number of the non-vehicle counting value is larger than 0, the current parking space is subjected to first counting processing, and if not, the current parking space is subjected to second counting processing.

The first counting process is as follows: counting the number of the vehicles is clear 0, and counting the number of the vehicles without the vehicles is added with 1; the second counting process is as follows: the no-vehicle counting value is cleared to be 0, and the vehicle counting value is added with 1. The initial value of the no-vehicle counting value and the vehicle counting value of the parking space is 0.

Further, the parking space state updating module 6 includes:

the parking space unknown state updating module 61 is used for updating the current parking space state to drive-in if the vehicle-containing count value of the current parking space is greater than a first threshold value when the current parking space state is unknown, updating the current parking space state to drive-out if the vehicle-free count value of the current parking space is greater than the first threshold value, and otherwise, not updating the current parking space state;

the parking space driving-in state updating module 62 is configured to count time when a license plate exists in a current parking space when the current parking space state is driving-in, update the current parking space state to occupied if the time when the license plate exists is greater than a second threshold, update the current parking space state to driving-out if a vehicle-free count value of the current parking space is greater than a first threshold, and otherwise, not update the current parking space state;

the parking space occupation state updating module 63 is configured to update the current parking space state to an outgoing state if the no-vehicle count value of the current parking space is greater than a first threshold value when the current parking space state is occupied, and not update the current parking space state if the no-vehicle count value of the current parking space is greater than the first threshold value;

the parking space exit state updating module 64 is used for counting the time when the current parking space state is exit, counting the time when no license plate exists in the current parking space, updating the current parking space state to be idle if the time when no license plate exists is greater than a second threshold, updating the current parking space state to be an entry state if the number of vehicles in the current parking space is greater than a first threshold, and otherwise, not updating the current parking space state;

and the parking space idle state updating module 65 is configured to update the current parking space state as entering if the current parking space state is idle and the vehicle-number count value of the current parking space is greater than the first threshold, otherwise, the current parking space state is not updated.

Wherein, the initial value of current parking stall state is unknown.

Parking stall state and license plate information output module 7 includes: and the parking space state and the corresponding license plate information of each parking space region in the current frame roadside video image are output.

While the foregoing is directed to the preferred embodiment of the present invention, and is not intended to limit the scope of the invention, it will be understood that the invention is not limited to the embodiments described herein, which are described to assist those skilled in the art in practicing the invention. Further modifications and improvements may readily occur to those skilled in the art without departing from the spirit and scope of the invention, and it is intended that the invention be limited only by the terms and scope of the appended claims, as including all alternatives and equivalents which may be included within the spirit and scope of the invention as defined by the appended claims.

Claims

1. A roadside parking space state identification method is characterized by comprising the following steps:

step seven, outputting the parking space states and license plate information of all parking spaces;

further, the fifth step includes:

and (3) processing the unallocated license plate, if the license plate is not allocated to the current parking space, performing the following processing: if the current parking space has no historical license plate information, processing according to the step of no historical license plate processing, otherwise, processing according to the step of historical license plate processing;

further, the second step includes:

a license plate to be matched output step, wherein the recognized license plate is used as a license plate to be matched, and the license plate to be matched and the corresponding position are output;

the third step includes:

a third matching step, matching the residual unallocated license plates with the non-parking space license plates, allocating the matched unallocated license plates to corresponding non-parking spaces, and establishing new non-parking space license plate information for the unmatched unallocated license plates;

the license plate matching is to compare license plate numbers one by one, if the license plate numbers are all the same, the license plate numbers are considered to be matched, otherwise, the license plate numbers are considered to be not matched;

further, the first counting process is: counting the number of the vehicles is clear 0, and counting the number of the vehicles without the vehicles is added with 1; the second counting process is as follows: counting the number of the non-vehicles by 0 and counting the number of the vehicles by 1; the initial value of the no-vehicle counting value and the vehicle counting value of the parking space is 0.

2. The method of claim 1, wherein the region of interest acquiring step comprises:

expanding the circumscribed rectangular region according to the upper boundary y of the circumscribed rectangular region₁And the lower boundary y is y₂X is left boundary₁And the right boundary x ═ x₂ObtainingThe upper boundary y of the region of interest₁-k₁h. Lower boundary y ═ y₂+k₂h. Left border x ═ x₁-k₃w, right boundary x ═ x₂+k₄w, wherein k₁、k₂、k₃、k₄If the boundary of the region of interest exceeds the boundary of the roadside video image, setting the exceeding boundary as the boundary of the corresponding roadside video image;

further, k is₁、k₂、k₃、k₄The value range of (A) is 0.1-0.6.

3. The method of claim 1, wherein the no history license plate processing step comprises: if the number of the non-vehicle counting values of the current parking space is larger than 0, performing first counting processing on the current parking space, and otherwise, performing second counting processing on the current parking space;

4. The method of claim 1, wherein the sixth step comprises:

5. The method of claim 4, wherein the first threshold value ranges from 1 to 120, and the second threshold value ranges from 5 to 40 seconds.

6. The utility model provides a roadside parking stall state recognition device which characterized in that, the device includes:

the parking space state and license plate information output module is used for outputting the parking space states and license plate information of all parking spaces;

further, the 1 occlusion parameter processing module includes:

and the unallocated license plate processing module is used for performing the following processing if the license plate is not allocated to the current parking space: if the current parking space has no historical license plate information, processing according to a no-historical license plate processing module, otherwise, processing according to a historical license plate processing module;

wherein, no historical license plate processing module includes: the system comprises a first counting module, a second counting module and a control module, wherein the first counting module is used for carrying out first counting processing on a current parking space if the number of the non-vehicle counting modules of the current parking space is more than 0, and otherwise, the second counting processing is carried out on the current parking space;

the historical license plate processing module comprises: the system is used for performing first counting processing on the current parking space if the previous parking space has no license plate information, and if the current parking space has no license plate information, the counting value is greater than 0, and otherwise, performing second counting processing on the current parking space; if the previous parking space has license plate information and the current parking space is matched with the license plate of the previous parking space, if the previous parking space is allocated to the license plate or the number of the current parking space is greater than 0, performing first counting processing on the current parking space, otherwise, performing second counting processing on the current parking space; if the number of the current parking space is larger than 0, performing first counting processing on the current parking space, otherwise, performing second counting processing on the current parking space;

the license plate output module to be matched is used for taking the identified license plate as a license plate to be matched and outputting the license plate to be matched and a corresponding position;

the license plate assignment module includes:

the third matching module is used for matching the residual unallocated license plates with the non-parking space license plates, allocating the matched unallocated license plates to the corresponding non-parking spaces, and establishing new non-parking space license plate information for the unmatched unallocated license plates;

7. The apparatus of claim 6, wherein the parking space status update module comprises: