CN102110376A - Roadside parking space detection device based on computer vision - Google Patents
Roadside parking space detection device based on computer vision Download PDFInfo
- Publication number
- CN102110376A CN102110376A CN 201110040674 CN201110040674A CN102110376A CN 102110376 A CN102110376 A CN 102110376A CN 201110040674 CN201110040674 CN 201110040674 CN 201110040674 A CN201110040674 A CN 201110040674A CN 102110376 A CN102110376 A CN 102110376A
- Authority
- CN
- China
- Prior art keywords
- parking stall
- sampled point
- parking
- vision sensor
- road
- 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.)
- Granted
Links
Images
Abstract
The invention discloses a roadside parking space detection device based on computer vision, comprising an omnidirectional vision sensor and a microprocessor, wherein the omnidirectional vision sensor is installed at the roadside with parking spaces; the microprocessor is used for analyzing the parking space state according to the data of a parking space video shot by the omnidirectional vision sensor; and the microprocessor comprises an omnidirectional image acquisition unit, a parking space customizing module, a parking space state detection module based on a sampling point model, and a parking space state information releasing module. The invention provides the roadside parking space detection device which is based on the computer vision, has the advantages of wide detection range, high detection precision, good detection instantaneity, high cost performance, strong environmental adaptation capability and good expandability and sustainable development, is convenient to implement and maintain and is convenient to realize on an embedded system.
Description
Technical field
The invention belongs to the application of optics, computer vision technique, mode identification technology, embedded system software and hardware technology and intelligent transport technology, especially a kind of parking position condition checkout gear based on computer vision at aspects such as the detection of the parking stall of road both sides, the management of parking charge and parking guidances.
Background technology
Current traffic problems have become global " city common fault ", and traffic congestion is the main performance of city " traffic illness "." cause of disease " of urban traffic blocking comes from multiple factor, and traffic congestion directly affects people's trip quality, particularly utilizes the people of vehicular traffic." parking difficulty " problem has become main " cause of disease " of urban traffic blocking at present.
Data shows that motor vehicle is in the obvious time that surpasses under the dynamic traffic state of time under the static traffic state, and both ratios are about 7: 1.Dynamic traffic provides the driveway space for vehicle, and static traffic provides the standage for vehicle.Advanced parking guidance system can improve the parking lot utilization factor, the environmental pollution that reduces vehicle owing to seek the parking stall and cause at the time of cruising on the road or in the parking lot and discharge tail gas and noise, improve whole traffic efficiency, improve the business operation condition in parking lot and the economic vitality of increase commerce area etc.
[i]
Parking guidance system generally four parts such as induces to form by parking information collection, information processing, information transmission and information.Wherein the parking information acquisition module provides information source for parking guidance system, and it is the key link that successfully realizes parking guidance system.The information acquiring technology on parking stall mainly contains following method at present: technology such as magnetic detection, ultrasound examination, infrared detection and Video Detection.
The inductive coil detecting device is traditional information detector, is present a kind of parking stall checkout equipment of consumption maximum in the world.Using method mainly is that the inductive coil sensor is embedded in detected parking stall subsurface, and parameter such as the caused voltage of coil changes when on the automatic detection parking stall car being arranged, thereby the perception parking stall takies situation.This method also has following shortcoming: 1) inductive coil must be imbedded under the ground, parking lot when mounted, therefore business as usual can to influence the parking lot when mounted, and when inductive coil breaks down, need excavate ground maintenance, maintenance workload and maintenance cost are big; 2) each parking stall all needs the installation and measuring device, and one-time investment is bigger, and many more can the bringing in parking stall communicated by letter and calculating pressure.
Ultrasonic detector is the situation that takies of being sent and being come the perception parking stall through the ultrasound wave of parking stall reflection by ultrasonic probe by accepting.Its principle of work can be divided into two kinds: propagation time difference method and Doppler method.Its weak point is affected by environment easily, people or thing that the probe below is passed through can produce reflection wave, cause flase drop easily, and the same each parking stall with the inductive coil detecting device all need the installation and measuring device, the initial stage input ratio is bigger, and many more can the bringing in parking stall communicated by letter and calculating pressure; This technology is in the context of detection of curb parking position, can occur the same problem of inductive coil detecting device if be embedded in below the ground, disapproves if be configured in the parking stall upper space.
Infrared detector is the suspension type detecting device with applications well prospect.The reflective detection technique of the general employing of this detecting device.Reflective detector probe is made up of an infraluminescence pipe and an infrared receiving tube, its principle of work is to produce modulating pulse by modulation pulse generator, through infrared probe radiation on the parking stall, when on the parking stall car being arranged, infrared pulse reflects from the parking stall, the receiving tube of being popped one's head in receives, through infrared detuner demodulation, by detection signal of trigger output after gating, amplification, rectification and the filtering.This detecting device has quick and precisely, the detectability of clear-cut.Its shortcoming is on-the-spot dust, the operate as normal that foreign material can influence detecting device.Above-mentioned three kinds of modes all belong on principle and detect or line detects.
Video Detection is a kind of technology in conjunction with video image and pattern-recognition.Compare with traditional detecting device, video parking information acquisition technique has the following advantages: 1) sensing range is wide, and the logical ccd video camera of a common Daepori can be monitored a plurality of parking stalls, and system can handle the video data that the logical ccd video camera of many Daeporis is gathered; 2) installation and maintenance are noiseless, because video camera generally is mounted in the top on parking stall, therefore installation and maintenance can not influence the business in parking lot, do not need excavation yet, destroy ground; 3) easy to maintenance, traditional inductive coil detecting device needs excavation ground to safeguard when damaging, and during video detecting device generation problem, can directly extract or repair facility, and has reduced maintenance cost; 4) visuality is good, realtime graphic can be transferred to the supvr in parking lot, realizes the monitoring function; 5) Video Detection belongs to the detection of face, and the quantity of information that is obtained is abundanter; 6) have good advance, extensibility, sustainable development etc.Therefore, video parking information acquisition technique is a kind of new trend of following parking stall detection technique development.
The Chinese invention patent application number is 200510127612.8 to disclose a kind of Roadside Parking course management system based on video technique, this system contains the ccd video camera of top, Roadside Parking field, extract the remote card reader of vehicle electric number plate, the video frequency car position detector that links to each other with ccd video camera; Detecting the parking stall state in this invention and be the gray-scale value that gray-scale value and vehicle by ground occupy the parking stall compares, this parking stall condition detection method robustness is lower, and the gray-scale value on ground is subjected to disturbing effects such as surround lighting, shade easily under the road environment especially out of doors;
The Chinese invention patent application number is that 200610002605.X discloses a kind of detecting system for vacancy of parking lots, comprises parking stall, ground marker character (2), is used to absorb wireless imaging sensor (1), central computer disposal system (3), parking stall status display system (4), parking guidance system (5) and the parking stall management software system (6) of parking stall state.Image sensor (1) receives the parking stall state of parking stall marker character (2), the treated central computer disposal system (3) that is transferred to, obtain the parking stall occupied information, be used for parking guidance system (5), parking stall status display system (4) is used for the supvr by parking stall management software system (6).Each parking stall is provided with a parking stall marker character (2), and it is made up of the figure of certain geometrical shape, and the image pretreatment module (8) of parking stall marker character (2) is housed on the image sensor (1), and the central computer disposal system is equipped with Target Identification Software on (3).The major defect of this invention is to draw a parking stall marker character on the ground, thereby because the road surface of road both sides is caused detecting failure through regular meeting by coverings such as rubbish, dust or snow;
All disclose the installation method of video camera in the invention in above-mentioned two inventions, if in general the video camera visual angle of installing is little, the also multipotency of video camera photographs 3～4 parking stalls so; If obtain wider visual angle, as covering space, 10 above parking stalls with a video camera, video camera need be installed in more than 10 meters the buildings outer wall or above the pole, this mounting means can be subjected to all-environment restriction, such as the blocking of road both sides trees, in appropriate place whether such buildings etc. is arranged, 10 meters costs with upper pright stanchion are very high, also brought problems such as safeguarding difficulty simultaneously;
Aspect the state-detection of parking stall, Funck (S.Funck, N.Mohler, W.Oertel.Determining car-park occupancy from single images[A] Proc.IEEE Intelligent Vehicles Symposium[C] .Parma, Italy, 2004:325-328.) and Mohd (Mohd.Yamani Idna, Idris Emran Mohd.Tamil.An Intelligent Parking Information System[A] .Proceedings of 3rd International Conference on Artificial Intelligence in Engineering and Technology[C], Sabah, MALAYSIA, 2006:22-24.) etc. people when using the background subtraction separating method in the parking stall is detected, at first make up a no car parking lot image as the reference background, with the image difference extraction prospect of new collection.This method does not need to consider how to obtain the parking stall earth background that vehicle takies, but this method need make up a no car parking lot image in earlier stage as the reference background, inconvenient in practice, and change or when occurring leaf, polybag etc. on the ground, parking lot and fly upward thing when ambient lighting brightness, because reference background does not have self-adaptation to cause foreground extraction inaccurate, has added a large amount of error messages in the prospect of extraction.Lin Sheng-Fuu (Lin Sheng-Fuu, Chen Yung-Yao, Liu Sung-Chieh.A vision-based parking lot management system[A] .Systems, Man and Cybernetics, 2006.SMC ' 06.IEEE International Conference on[C] .New York, 2006:2897-2902.) etc. the people carry out the parking stall in conjunction with reference background model and Gaussian Background model adaptation update method and detect, to solve reference background model background adaptive updates problem, this model utilizes the adaptive updates method in the Gaussian Background model to carry out context update to zone, empty wagons position, and earth background carried out context update near the zone, parking stall that has vehicle to take was adopted.But this background model update method must be close with the ground color in two zones is prerequisite, otherwise this method lost efficacy, and this method must will obtain the parking stall earlier and take situation and could determine to adopt which kind of mode to carry out context update, thereby in case the failure error takes place to detect will accumulate.M.Durus (M.Durus, A.Ercil.Robust vehicle detection algorithm[J] .Signal Processing and Communications Applications[C] .Eskisehir, 2007:1-4.) adopt the vehicle tracking method to carry out the parking stall to detect, extract the moving vehicle track and detect the parking stall and take situation by following the tracks of vehicle.This method need not to make up reference background, but this method is calculated more complicated, and relatively more difficult aspect extraction microinching track of vehicle.
The other researchist takies situation by directly utilizing vehicle object or parking stall characteristics of objects to discern the parking stall.Ping.Y (Ping, Y.Jiang, D.L.Wei.Parking stall detection based on Fisher discrimination of multiple static imaging feature[A] .Proceedings of the First International Symposium on Test Automation ﹠amp; Instrumentation[C] .Beijing, 2006:319-323.) adopt fisher carrying out image threshold segmentation method to carry out the parking stall and detect.This detection model calculates simple and can overcome the influence of little target to image segmentation, but this detection model can not effectively solve the shadow problem that exists in the image, and vehicle class is various, utilizes single threshold value to classify and is difficult to guarantee the parking stall accuracy of detection.N.Srinivasa (N.Srinivasa.Method and apparatus for the surveillance of objects in images[P] .United States Patent No.0235327,2003.) and D.B.L.Bong (D.B.L.Bong, K.C.Ting, K.C.Lai.Integrated Approach in the Design of Car Park Occupancy Information System (COINS) [J] .International Journal of Computer Science, 2008,35 (1): 7-14.) adopt the edge that extracts in the parking stall to take judging characteristic as the parking stall, this detection model strong interference immunity can adapt to various illumination brightness variable effects, but when this detection model comprised too many background edge information in running into parking stall regional background more complicated or parking stall, prospect or background problems can't be judged in the edge that can produce extraction.Consider that vehicle shape is well-regulated, S.Gupte (Gupte, S.Masoud, O., Martin, R.F.K.Detection and classification of vehicles[J] .Intelligent Transportation Systems, IEEE Transactions on, 2002,3 (1): 37-47.) whether exist rectangle to take judging characteristic in the analysis image as the parking stall, but this detection model need relate to a large amount of calculating, and computation complexity is than higher.K.Iwasa (K IWASA, T TANAKA, Y SAGAWA.Automatic Generation of Template Images for Detecting Vehicles in Parking Lots [J] .Transactions of the Institute of Electrical Engineers of Japan.2007,127 (3): 338-344.) adopt template matching method to carry out the parking stall and detect, utilize many parking lot image matches to obtain the parking lot image of a no car as the reference background, cut apart the zone, parking stall by the parking stall white line in the detected image, generate the parking stall template automatically.The parking stall template that this detection model utilization generates automatically detects the advantage with strong interference immunity, but generate the algorithm more complicated of parking stall template automatically, because the parking stall white line is not obvious or problem such as block, cause the parking stall template of automatic generation and actual parking stall area differentiation bigger on the other hand.Siming Liu (Siming Liu.Robust vehicle detection from parking lot images[J] .Boston University, Dept.of Electrical and Computer Engineering, 2005:3-8.) adopt the quaternary tree partitioning algorithm to carry out the parking stall to detect, the result of cutting apart according to quaternary tree adds up the cut zone number that comprises in the parking stall and detects the parking stall and take situation.
Though above-mentioned achievement in research can be carried out the parking stall status detection under specific situation,, obviously do not fit under embedded system environment and use because the computation complexity height needs a large amount of computational resources and storage resources in computation process;
In order under different environment, all to obtain high measurement accuracy and high reliability, necessary employing strong interference immunity, the parking stall Video Detection algorithm that robustness is high.Colour TV camera is used quite general at present, how to use color information effectively and be used as the parking stall status detection, this detection can suitably improve accuracy of detection can effectively utilize computational resource and storage resources again (in some original detection algorithms simultaneously, convert existing color information to gray value information, do not utilize these resources well); In addition, some background modeling technology have consumed very a large amount of calculating and storage resources in above-mentioned achievement in research, in the detection on parking stall, each parking stall has individual or thousands of the pixels of hundreds of basically, if only several representative pixels are wherein carried out background modeling, can reduce the computational resource and the storage resources of embedded system effectively;
Therefore a kind of design proposal of outstanding curb parking level detecting apparatus must be followed 6 principles: 1) reliability or robustness height; 2) the environment self-adaption ability is strong; 3) accuracy of detection height or false detection rate are low; 4) convenient for installation and maintenance; 5) ratio of performance to price height; 6) extensibility, sustainable development are good.
Summary of the invention
For the limitation of the detection that overcomes existing curb parking level detecting apparatus big, implement investment and maintenance cost height, the detection means of contact is unfriendly to road, the Video Detection algorithm is difficult to implement on embedded system, false detection rate is than higher, adaptive capacity to environment is lower, the ratio of performance to price also is difficult to by deficiencies such as market acceptance, the invention provides that a kind of to have sensing range wide, the accuracy of detection height, the detection real-time is good, implement easy to maintenance, ratio of performance to price height, adaptive capacity to environment is strong, extensibility, the curb parking level detecting apparatus that sustainable development is good based on computer vision.
The technical solution adopted for the present invention to solve the technical problems is:
A kind of curb parking level detecting apparatus based on computer vision, comprise the road roadside that is installed in the parking stall panoramic vision sensor, be used for analyzing the microprocessor of parking stall state according to the captured parking stall video data of panoramic vision sensor, described panoramic vision sensor is connected with described microprocessor by video interface, and described panoramic vision sensor and described microprocessor are fixed in the vertical rod in road roadside; The field range of described panoramic vision sensor comprises that described microprocessor comprises along the parking stall in the parking place of road direction:
The panoramic picture acquisition module is used to obtain initialization information and video image, comprises system initialization unit and image acquisition unit; In the described system initialization unit, threshold value achievement data, parking stall customization data and sampled point spatial positional information are read in the dynamic storage cell, call in order in the subsequent processes; In the described image acquisition unit, read from panoramic vision sensor and pass the video image information come and video image information is kept at the dynamic storage cell;
The parking stall customized module, be used for according to the parking stall virtual parking stall that match of the parking stall of delimiting on the road under customization on the panoramic picture that obtains and actual parking situation, behind the good virtual parking stall of customization, need to customize again the detection sampled point, customize the detection sampled point of setting quantity among the present invention in each virtual parking stall, the information of sampled point is kept in the storage unit in virtual parking space information that customization is good and the virtual parking stall;
Based on the parking stall state detection module of sampled point model, be used to detect in the virtual parking stall by the vehicle possession state;
Parking stall status information release module, be used for all the parking stall states in the parking place are sent to the parking guidance device, described parking guidance device occupies situation is dynamically issued the parking stall, parking place in somewhere by various distributing devices the situation of occupying according to the parking stall in the parking place.
As preferred a kind of scheme: described panoramic vision sensor becomes the minute surface and the camera lens of angles just to be constituted towards the video camera of minute surface by two; Angle between two minute surfaces is 180 °-2 γ, and two minute surfaces are W at the width value on the front elevation, the height value on side view is R, and the width value W of two minute surfaces and height value R are positioned at the imaging scope of video camera; On side view, the central shaft of described video camera becomes the η angle with the central shaft of described vertical rod, and minute surface becomes the ε angle with surface level direction road side; On front elevation, the angle of minute surface and described vertical rod is 90 °-γ, and the central shaft of video camera and the central axes of vertical rod, focus of camera are f;
The setting height(from bottom) of panoramic vision sensor is H, and the road surface is L along the visual range of road direction, 180 °-2 γ of angle between two minute surfaces, and formula (1) is the relation of H, L value and γ,
In the formula, γ represents the angle of minute surface and surface level, L be panoramic vision sensor along the visual length on the surface level direction road, H is the setting height(from bottom) of panoramic vision sensor,
Maximum visual angle for video camera.
Further, the maximum visual angle of described video camera
It is 45 °, the setting height(from bottom) H of panoramic vision sensor is 3 meters, panoramic vision sensor be 100 meters along the visual length L on the road direction, the angle γ that tries to achieve minute surface and surface level by formula (1) is 32 °, the length of minute surface is greater than W/2 * cos (γ), and the width of every minute surface is greater than R/cos (a-η).
Further, in the described parking stall state-detection based on the sampled point model, detect the degree of the sampled point covering in virtual parking stall and judge whether this parking stall is occupied, described parking stall state-detection calculation process Sa～Sh is as follows:
In the calculation procedure of Sa, according to from panoramic vision sensor in the position on the road to the pairing road of each sampled point the distance of physical location come setting threshold TH1, threshold value TH1 is provided with as criterion in the back to the binary conversion treatment of sampled point the time;
In the calculation procedure of Sb, the panoramic picture that tn is obtained constantly is processed into the sampled point image by the pairing pixel of sampled point, and the gray-scale value of pairing each pixel of sampled point on the sampled point image is represented with 8 bit data;
In the calculation procedure of Sc, calculate the difference between benchmark gray level image and the described sampled point image, obtain the difference image of two width of cloth images;
In the calculation procedure of Sd, carry out the background modeling of benchmark gray level image, bring in constant renewal in tn benchmark gray level image B constantly with formula (2)
nObtain tn+1 benchmark gray level image B constantly
N+1:
B
n+1＝B
n+φ×(X
n-B
n) (2)
In the formula, X
nBe the gray-scale value of each sampled point in the t n sampled images constantly, B
nBe the gray-scale value of each sampled point of t n benchmark gray level image constantly, B
N+1Be the gray-scale value of each sampled point of t n+1 benchmark gray level image constantly, φ is the very little coefficient of a numerical value;
When calculating, at first calculate (the X of each sampled point in the parking stall
n-B
n) value, getting the absolute value of its value then | X
n-B
n|, if this absolute value | X
n-B
n| greater than the threshold value TH2 B of this sampled point simultaneously of regulation
nNearest non-of value and this sampled point exists the absolute value of gray-scale value of sampled point less than the threshold value TH3 of regulation, just be judged to be foreground object and entered on this sampled point, the renewal of this sampled point at this moment just with the nearest non-gray-scale value of sampled point that exists of this sampled point as B
N+1The background modeling of all the other sampled points all by formula (2) upgrades processing;
In Se and Sf calculation procedure, be used in each the threshold value TH1 that sets in the Sa step and carry out binary conversion treatment, all sampled points will be divided into " 0 " or " 1 " two states, have foreground object to exist on this sampled point of the expression of " 1 ", not have foreground object on this sampled point of the expression of " 0 ";
In the calculation procedure of Sg, adopt auto model to eliminate the misinterpretation of these sampled points; Specific algorithm is: exist sampled point around a non-situation that has sampled point is arranged, should non-ly exist sampled point according to auto model as misinterpretation, should non-ly exist sampled point to be modified to and have sampled point; Not existing around under the situation of sampled point has to have a sampled point; Exist the situation of foreground object on the described sampled point that exists sampled point to be illustrated in the parking stall, do not exist the situation of foreground object on the described non-sampled point that exists sampled point to be illustrated in the parking stall;
In the calculation procedure of Sh, all the sampling number purposes that exist in sampled point number and the parking stall in the calculation procedure according to above-mentioned Sg on resultant certain parking stall judge recently whether this parking stall occupied, computing method as shown in Equation (3),
Rate＝(T
ec/T
c)×100 (3)
In the formula, T
EcBe the statistical value that has sampled point in certain parking stall, T
cBe the statistical value of all sampled points in certain parking stall, Rate is the space hold ratio in certain parking stall;
If calculated value Rate, is judged to be this parking stall more than or equal to threshold value TH4 and is taken by vehicle.
The employing circular arc minute surface of described panoramic vision sensor, described video camera is f to the distance of circular arc minute surface, the camera field of view scope is W * R, the panoramic vision sensor setting height(from bottom) is H, is L along the visual length on the road direction, and the radius-of-curvature of circular arc minute surface is W, the central angle of circular arc is 60 °, the arc length of circular arc minute surface is π * W/3, ε be the circular arc minute surface on side view with the angle of surface level, η is the angle between video camera and the vertical rod.
The naming method of described virtual parking stall: all parking stalls are with one group of array representation, PS
i* (n), n with in the face of the shooting direction of panoramic vision sensor from left to right with 1～10 series arrangement, PS
iBe illustrated on this road from left to right i panoramic vision sensor, * option have only two, N and A, wherein N represents the road side near panoramic vision sensor, A represents the road side away from panoramic vision sensor.
In the status information release module of described parking stall, all the parking stall states in the parking place are sent to the management of parking charge system, described the management of parking charge system chargeed according to the parking stall time that a certain car occupies.
Described parking guidance device is that road traffic is induced display board.
Or: described parking guidance device is the webpage of intelligent transportation, and the mode that the user can browsing page is understood the situation that takies on parking stall on the road of somewhere.
Described parking guidance device is the various service units of mobile Internet, and the user understands the situation that takies on parking stall on the road of somewhere in real time by short message mode.
Beneficial effect of the present invention mainly shows: video information on a large scale that 1, can the whole road of real-time collecting, and it is wide to have sensing range, can the parking stall state about 20 be detected; 2, installation and maintenance are noiseless, because video detector is installed on the pedestrian road of road often, therefore installing and safeguarding not to influence the current of road, does not need excavation yet, destroys the road surface; 3, adopt the mode of sampled point modeling to make calculated amount and memory space will reduce hundred times, help on embedded system, realizing than original technology; 4, detecting reliability, accuracy height have self study and intelligent function; 5, can be connected like toll administration unit, system for traffic guiding with various advanced persons' traffic control system by network or communication interface, realize the more intelligent traffic management function.
Description of drawings
Fig. 1 is the detection synoptic diagram based on the curb parking position condition checkout gear of computer vision;
Fig. 2 is front elevation and the side view that is used for the panoramic vision sensor of curb parking position condition checkout gear;
Fig. 3 is the image synoptic diagram of the captured curb parking position state of panoramic vision sensor;
Fig. 4 is the physical size size and the sampled point customization synoptic diagram on parking stall;
Fig. 5 (a) is the situation synoptic diagram that sampled point is blocked when vehicle parking is on the parking stall, and Fig. 5 (b) detects the key diagram that has sampled point when having vehicle parking on the parking stall with Fig. 5 (a);
Fig. 6 is the another kind of schematic diagram that is used for the panoramic vision sensor of curb parking position condition checkout gear;
Fig. 7 is for judging the process flow diagram of parking stall state;
Fig. 8 is the synoptic diagram that customizes the sampled point in parking stall and the parking stall on panoramic picture;
Fig. 9 detects the synoptic diagram of the sampled point in the parking stall for park panoramic picture under the situation of Fig. 3;
Figure 10 is for considering the synoptic diagram on customizing virtual parking stall under the situation of the visual angle of panoramic vision;
Figure 11 is the system's pie graph based on the curb parking position condition checkout gear of computer vision;
Figure 12 (a) is for removing the non-synoptic diagram that has the interference of sampled point in some parking stalls; Figure 12 (b) is for removing the actual non-synoptic diagram that has the interference of sampled point in some parking stalls.
Embodiment
Below in conjunction with accompanying drawing the present invention is further described.
With reference to Fig. 1～5, Fig. 7～12, a kind of curb parking level detecting apparatus based on computer vision, comprise the road roadside that is installed in the parking stall panoramic vision sensor, be used for analyzing the microprocessor 3 of parking stall state according to the captured parking stall video data of panoramic vision sensor, described panoramic vision sensor is connected with described microprocessor 3 by video interface, described panoramic vision sensor and described microprocessor 3 are fixed in the vertical rod 4 in road roadside, as shown in Figure 1; The field range of described panoramic vision sensor is 10～15 parking stalls along road direction, basic consistent aspect width with the width of road, comprise embedded system hardware and built-in system software in the described microprocessor, described embedded system hardware comprises: CPU, video interface, storage unit, input block, image storage unit, RAM, ROM, display unit and delivery unit, annexation between them as shown in figure 11, CPU reads necessary programs from ROM, RAM is as the perform region of program, the implementation program, implementation by program, from image storage unit, read the necessary image data, calculate various related datas according to these view data, finally calculate the testing result of road parking stall state condition, the various testing results that program calculates after carrying out are kept in the storage unit, simultaneously with the export to communication unit of testing result by delivery unit, the output of communication unit has two, one sends to the toll administration unit, another sends to the parking guidance unit, input block is used to import various setting values and threshold value, a kind of man-machine interaction means that adopted in the time of as these data of change, display unit are used for the affirmation means when input block is set various threshold values and customization parking stall and sampled point; Built-in system software moves calculating in described CPU; Described built-in system software comprises:
The panoramic picture acquiring unit is used to obtain initialization information and video image, comprises system initialization module and image collection module;
System initialization module is used for threshold value achievement data, parking stall customization data and sampled point spatial positional information are read into dynamic storage cell, calls in order in the subsequent processes;
Image collection module is used for reading from panoramic vision sensor and passes the video image information of coming and video image information is kept at dynamic storage cell;
The parking stall customized module, be used for according to the parking stall virtual parking stall that match of the parking stall of delimiting on the road under customization on the panoramic picture that obtains and actual parking situation, on whole curb parking position owing to the reason at the visual angle of panoramic vision sensor, vehicle has certain height, to be subjected to the influence of blocking that parks cars on the previous parking stall from panoramic vision sensor distance parking stall far away more, as shown in figure 10; Will be during the customization parking stall according to customizing from the panoramic vision sensor distance, specific practice is that the stop line on the captured image of panoramic vision sensor is added a visual angle correction, second the parking stall stop line that begins from the video camera vertical rod as shown in figure 10 increases by 25%, the 4th parking stall of stop line increase, 12%, the 3rd parking stall, 40%, the 5th parking stall of stop line increase stop line increase by 80%, and the parking stall with these customizations is called virtual parking stall in the present invention; State on the detection parking stall is only limited to the image section in the virtual parking stall; The title on parking stall is identical on the naming method of virtual parking stall and the real road; If the parking stall is not named on the real road, then to name by the method that proposes among the present invention, all parking stalls describe PS with one group of array representation with Fig. 8
i* (n), n with in the face of the shooting direction of panoramic vision sensor from left to right with 1～10 series arrangement, PS
iBe illustrated on this road from left to right i panoramic vision sensor, * option have only two, N and A, wherein N represents the road side near panoramic vision sensor, A represents the road side away from panoramic vision sensor; Behind the good virtual parking stall of customization, need to customize again the detection sampled point, in each virtual parking stall, customize 24 uniform detection sampled points among the present invention, as shown in Figure 4; The information of sampled point is kept in the storage unit in virtual parking space information that customization is good and the virtual parking stall;
Based on the parking stall state detection module of sampled point model, be used to detect in the virtual parking stall by the vehicle possession state;
Parking stall status information release module, be used for all the parking stall states in the parking lot are sent to parking guidance unit or toll administration unit by delivery unit and communication unit, described parking guidance unit occupies situation is dynamically issued the parking stall, parking lot in somewhere by various distributing devices the situation of occupying according to the parking stall in the parking lot; Described toll administration unit chargeed according to the parking stall time that a certain car occupies;
A kind of form of described distributing device is that road traffic is induced display board; Or: the another kind of form of described distributing device is the webpage of intelligent transportation, and the mode that the user can browsing page is understood the situation that takies on parking stall on the road of somewhere; Or be: the another kind of form of described distributing device is the various services of mobile Internet, and the user can understand the situation that takies on parking stall on the road of somewhere in real time by modes such as notes;
Described panoramic vision sensor just is made of towards the video camera 1 of minute surface two angled minute surfaces 2 and camera lens, as shown in Figure 2; Fig. 2 (a) is a front elevation, promptly observes the view of panoramic vision sensor from road; Fig. 2 (b) is a side view, promptly observes the view of panoramic vision sensor from a side of road; Angle between two minute surfaces 2 is 180 °-2 γ, and two minute surfaces 2 are W at the width value on the front elevation, the height value on side view is R, and the width value W of two minute surfaces 2 and height value R are just in time in the imaging scope of video camera; On side view, the central shaft of video camera becomes the η angle with the central shaft of vertical rod, and minute surface 2 becomes the ε angle with surface level to road side; On front elevation, minute surface 2 is 90 °-γ with the angle of vertical rod, the central shaft of video camera and the central axes of vertical rod, and the focal length of video camera 1 is f; The captured panoramic picture of described panoramic vision sensor as shown in Figure 3; Fig. 4 is the big logotype in the parking stall of parking lay-by, black box among Fig. 3 is represented parking stall one by one, road both sides, design according to as shown in Figure 2 panoramic vision sensor, shape from video camera two parking stalls the most nearby is basic identical with the shape on actual parking stall, along with the parking stall size shape of distance increasing on the panoramic imagery plane on video camera and parking stall dwindled along road direction is proportional, its proportionate relationship is relevant with the angle between the minute surface 2;
The design of panoramic vision sensor is described, the setting height(from bottom) of panoramic vision sensor is H, and the road surface is L along the visual range of road direction, designs 180 °-2 γ of angle between two minute surfaces 2, and formula (1) is the relation of H, L value and γ,
In the formula, γ represents the angle of minute surface and surface level, L be panoramic vision sensor along the visual length on the road direction, H is the setting height(from bottom) of panoramic vision sensor,
Maximum visual angle for video camera;
If the maximum visual angle of video camera
It is 45 °, the setting height(from bottom) H of panoramic vision sensor is 3 meters, panoramic vision sensor be 100 meters (being equivalent to 16 parking stalls) along the visual length L on the road direction, the angle γ that tries to achieve minute surface and surface level by formula (1) is 32 °, the length of minute surface is slightly larger than W/2 * cos (γ), the width that designs every minute surface be slightly larger than R/cos (ε-η), ε be the circular arc minute surface on side view with the angle of surface level, η is the angle between video camera and the vertical rod;
To determine from the distance of video camera vertical rod according to the parking stall with the angle η of vertical rod and the angle ε of minute surface 2 and surface level about video camera in the side view shown in Fig. 2 (b) 1, generally satisfy the requirement of obtaining the parking stall image with the on-the-spot mode of adjusting angle; The width segments that is noted that video camera 1 photographic images is along road direction, and the height of image partly is perpendicular to road direction, as shown in Figure 3;
Described parking stall state-detection based on the sampled point model, be used for detecting the degree that the sampled point in virtual parking stall covers and judge whether this parking stall is occupied, as shown in Figure 4,24 sampled points are arranged in a virtual parking stall, sampled point is represented with " ● ", sampled point is kept in the storage unit with the name of virtual parking stall with the two-dimensional array form, as for PS
iPreserve the pixel coordinate value on the panoramic picture of 24 sampled points in the parking stall of N (5), all included 24 sampled points in the parking stall of each customization; If under 20 parking stall situations shown in Figure 8, then include 480 sampled points; When vehicle had taken the parking stall, if at this moment there are 12 sampled points to be capped, we were called the sampled point that is capped and have sampled point, as shown in Figure 5; Judge recently by what detect the sum that has sampling number and sampled point in the parking stall whether the parking stall is occupied among the present invention, can effectively get rid of some interference in the testing environment like this, as the pedestrian parking stall of passing by, flying upward thing as leaf, polybag etc. just in time drops in the parking stall, these only disturb can reduce by 1 maximum several sampled point that exist, and these sampled points also are isolated the existence; This detection method also is the detection that belongs to face, among the present invention with the parking stall as a detection faces that constitutes by 24 sampled points, simultaneously with the vehicle in the parking stall also as a detection faces; Therefore need only for background modeling 24 sampled points in the parking stall are carried out modeling, with need in original technology that all pixels in the parking stall are carried out modeling, its calculated amount and memory consumption have reduced hundred times approximately, have also eliminated the interference that several pixel produced simultaneously; Below we further specify detection method of the present invention with regard to parking stall state-detection calculation process shown in Figure 7;
In the calculation procedure of Sa, according to from panoramic vision sensor in the position on the road to the pairing road of each sampled point the distance of physical location come setting threshold TH1, this is to consider the far and near influence to threshold value TH1 of distance, in general, threshold value TH1 reduces from the near to the remote gradually according to image-forming range, this be because the difference of the gray-scale value between road background and the vehicle prospect along with the image-forming range increase can reduce; Be provided with as criterion for the time about threshold value TH1 in the back to the binary conversion treatment of sampled point;
In the calculation procedure of Sb, the panoramic picture that tn is obtained constantly is processed into the sampled point image by the pairing pixel of sampled point, Fig. 8 is the example of sampled point panoramic picture, and the gray-scale value of pairing each pixel of sampled point on the sampled point image is represented with 8 bit data;
In the calculation procedure of Sc, calculate the difference between benchmark gray level image and the described sampled point image, obtain the difference image of two width of cloth images, the purpose of subtracting each other is in order to obtain the sampled point that foreground object comprised on the parking stall, the result who subtracts each other is that the sampled point with identical gray-scale value on the sampling point position identical in two width of cloth images is cut, kept the sampled point that has different gray-scale values on the identical sampling point position, we are called prediction with this sampled point and have sampled point, perhaps are called prediction foreground object sampled point; Further we describe with accompanying drawing 4, accompanying drawing 5, accompanying drawing 4 is calculating benchmark gray level images of some parking stalls, accompanying drawing 5 (a) is the sampled point image, and accompanying drawing 4 deducts the result of accompanying drawing 5 (a) shown in accompanying drawing 5 (b), has only kept in certain parking stall on difference result figure and has had sampled point; Described benchmark gray level image, desirable state is at the sampled point image that does not have on the parking stall under the vehicle condition, but because the variation of weather and the sampled point gray-scale value of the passing benchmark gray level image of time also can change, this belongs to the background modeling problem, we set forth background modeling in next calculation procedure, the processing of promptly described benchmark gray level image;
In the calculation procedure of Sd, mainly carry out the background modeling of benchmark gray level image, bring in constant renewal in tn benchmark gray level image B constantly with formula (2) among the present invention
nObtain t n+1 benchmark gray level image B constantly
N+1
B
n+1＝B
n+φ×(X
n-B
n) (2)
In the formula, X
nBe the gray-scale value of each sampled point in the t n sampled images constantly, B
nBe the gray-scale value of each sampled point of t n benchmark gray level image constantly, B
N+1Be the gray-scale value of each sampled point of t n+1 benchmark gray level image constantly, φ is the very little coefficient of a numerical value, as long as be used for adapting to the influence of the variation of Changes in weather, passage of time and sunlight to the gray-scale value of each sampled point of benchmark gray level image;
When calculating, at first calculate (the X of each sampled point in the parking stall
n-B
n) value, getting the absolute value of its value then | X
n-B
n|, if this absolute value | X
n-B
n| greater than the threshold value TH2 B of this sampled point simultaneously of regulation
nNearest non-of value and this sampled point exists the absolute value of gray-scale value of sampled point less than the threshold value TH3 of regulation, just be judged to be foreground object and entered on this sampled point, the renewal of this sampled point at this moment just with the nearest non-gray-scale value of sampled point that exists of this sampled point as B
N+1The background modeling of all the other sampled points all by formula (2) upgrades processing;
In Se and Sf calculation procedure, be used in each the threshold value TH1 that sets in the Sa step and carry out binary conversion treatment, this result, all sampled points will be divided into " 0 " or " 1 " two states, there is foreground object to exist on this sampled point of the expression of " 1 ", do not have foreground object on this sampled point of the expression of " 0 "; But above-mentioned calculating also unavoidably can exist sampled point erroneous judgement to be decided to be to have sampled point with non-, equally also can will exist the sampled point mistake to be identified as the non-sampled point that exists, in order to eliminate these misinterpretations, adopt auto model to eliminate the misinterpretation of these sampled points among the present invention;
In the calculation procedure of Sg, adopt auto model to eliminate the misinterpretation of these sampled points; From imaging plane, if there is vehicle to occupy the place of parking stall, having sampled point should be to be linked to be a face, if in the face of a rectangle, exist the non-sampled point that exists, it just might be the misinterpretation in the above-mentioned processing procedure, color such as certain position of vehicle and ground is more approaching, makes the residing sampled point in this position be used as the non-sampled point that exists; Just in time be to drop on some sampled points such as leaf is arranged equally, be judged as and have sampled point, and do not have the continuous sampled point that exists near this sampled point; Misinterpretation for above-mentioned several situation sampled points can be eliminated by auto model; Shown in accompanying drawing 12 (a), among this figure be exist sampled point around a non-situation that has sampled point is arranged, should non-ly exist sampled point according to auto model as misinterpretation, should non-ly exist sampled point to be modified to and have sampled point; Shown in accompanying drawing 12 (b), be not exist around under the situation of sampled point to have to have a sampled point among this figure, exist sampled point as misinterpretation this according to auto model, and it is modified to the non-sampled point that exists; Can eliminate some misinterpretations by the above-mentioned correction based on auto model, that has eliminated isolated existence basically exists sampled point or the non-sampled point that exists; Exist the situation of foreground object on the described sampled point that exists sampled point to be illustrated in the parking stall, do not exist the situation of foreground object on the described non-sampled point that exists sampled point to be illustrated in the parking stall;
In the calculation procedure of Sh, all the sampling number purposes that exist in sampled point number and the parking stall in the calculation procedure according to above-mentioned Sg on resultant certain parking stall judge recently whether this parking stall occupied, computing method as shown in Equation (3),
Rate＝(T
ec/T
c)×100 (3)
In the formula, T
EcBe the statistical value that has sampled point in certain parking stall, T
cBe the statistical value of all sampled points in certain parking stall, Rate is the space hold ratio in certain parking stall;
If calculated value Rate is more than or equal to threshold value TH4, the TH4 value is generally 40, just is judged to be this parking stall and is taken by vehicle; Fig. 9 is the result of the parking stall state-detection under the parking stall parking situation among Fig. 3.
With reference to Fig. 1～4, Fig. 6～12, in the present embodiment, the employing of panoramic vision sensor circular arc minute surface shown in Figure 6 replaces two plate plane minute surfaces shown in Figure 2; Is f for known video camera 1 to the distance of circular arc minute surface, video camera 1 field range is W * R, the panoramic vision sensor setting height(from bottom) is H, along the visual length on the road direction is L, and the radius-of-curvature of circular arc minute surface is W, and the central angle of circular arc is 60 °, the arc length of circular arc minute surface is π * W/3, the width of circular arc minute surface be slightly larger than R/cos (ε-η), ε be the circular arc minute surface on side view with the angle of surface level, η is the angle between video camera and the vertical rod.
Other schemes of present embodiment are all identical with embodiment 1.
Claims (10)
1. curb parking level detecting apparatus based on computer vision, it is characterized in that: comprise the road roadside that is installed in the parking stall panoramic vision sensor, be used for analyzing the microprocessor of parking stall state according to the captured parking stall video data of panoramic vision sensor, described panoramic vision sensor is connected with described microprocessor by video interface, and described panoramic vision sensor and described microprocessor are fixed in the vertical rod in road roadside; The field range of described panoramic vision sensor comprises that described microprocessor comprises along the parking stall in the parking place of road direction:
The panoramic picture acquisition module is used to obtain initialization information and video image, comprises system initialization unit and image acquisition unit; In the described system initialization unit, threshold value achievement data, parking stall customization data and sampled point spatial positional information are read in the dynamic storage cell, call in order in the subsequent processes; In the described image acquisition unit, read from panoramic vision sensor and pass the video image information come and video image information is kept at the dynamic storage cell;
The parking stall customized module, be used for according to the parking stall virtual parking stall that match of the parking stall of delimiting on the road under customization on the panoramic picture that obtains and actual parking situation, behind the good virtual parking stall of customization, need to customize again the detection sampled point, customize the detection sampled point of setting quantity among the present invention in each virtual parking stall, the information of sampled point is kept in the storage unit in virtual parking space information that customization is good and the virtual parking stall;
Based on the parking stall state detection module of sampled point model, be used to detect in the virtual parking stall by the vehicle possession state;
Parking stall status information release module, be used for all the parking stall states in the parking place are sent to the parking guidance device, described parking guidance device occupies situation is dynamically issued the parking stall, parking place in somewhere by various distributing devices the situation of occupying according to the parking stall in the parking place.
2. the curb parking level detecting apparatus based on computer vision as claimed in claim 1 is characterized in that: described panoramic vision sensor becomes the minute surface and the camera lens of angle just to be constituted towards the video camera of minute surface by two; Angle between two minute surfaces is 180 °-2 γ, and two minute surfaces are W at the width value on the front elevation, the height value on side view is R, and the width value W of two minute surfaces and height value R are positioned at the imaging scope of video camera; On side view, the central shaft of described video camera becomes the η angle with the central shaft of described vertical rod, and minute surface becomes the ε angle with surface level direction road side; On front elevation, the angle of minute surface and described vertical rod is 90 °-γ, and the central shaft of video camera and the central axes of vertical rod, focus of camera are f;
The setting height(from bottom) of panoramic vision sensor is H, and the road surface is L along the visual range of road direction, 180 °-2 γ of angle between two minute surfaces, and formula (1) is the relation of H, L value and γ,
In the formula, γ represents the angle of minute surface and surface level, L be panoramic vision sensor along the visual length on the surface level direction road, H is the setting height(from bottom) of panoramic vision sensor,
Maximum visual angle for video camera.
3. the curb parking level detecting apparatus based on computer vision as claimed in claim 2 is characterized in that: the maximum visual angle of described video camera
It is 45 °, the setting height(from bottom) H of panoramic vision sensor is 3 meters, panoramic vision sensor be 100 meters along the visual length L on the road direction, the angle γ that tries to achieve minute surface and surface level by formula (1) is 32 °, the length of minute surface is greater than W/2 * cos (γ), and the width of every minute surface is greater than R/cos (ε-η).
4. as the described curb parking level detecting apparatus of one of claim 1～3 based on computer vision, it is characterized in that: in the described parking stall state-detection based on the sampled point model, the degree that the sampled point of detection in virtual parking stall covers judges whether this parking stall is occupied, and described parking stall state-detection calculation process Sa～Sh is as follows:
In the calculation procedure of Sa, according to from panoramic vision sensor in the position on the road to the pairing road of each sampled point the distance of physical location come setting threshold TH1, threshold value TH1 is provided with as criterion in the back to the binary conversion treatment of sampled point the time;
In the calculation procedure of Sb, the panoramic picture that tn is obtained constantly is processed into the sampled point image by the pairing pixel of sampled point, and the gray-scale value of pairing each pixel of sampled point on the sampled point image is represented with 8 bit data;
In the calculation procedure of Sc, calculate the difference between benchmark gray level image and the described sampled point image, obtain the difference image of two width of cloth images;
In the calculation procedure of Sd, carry out the background modeling of benchmark gray level image, bring in constant renewal in tn benchmark gray level image B constantly with formula (2)
nObtain tn+1 benchmark gray level image B constantly
N+1:
B
n+1＝B
n+φ×(X
n-B
n) (2)
In the formula, X
nBe the gray-scale value of each sampled point in the t n sampled images constantly, B
nBe the gray-scale value of each sampled point of t n benchmark gray level image constantly, B
N+1Be the gray-scale value of each sampled point of t n+1 benchmark gray level image constantly, φ is the very little coefficient of a numerical value;
When calculating, at first calculate (the X of each sampled point in the parking stall
n-B
n) value, getting the absolute value of its value then | X
n-B
n|, if this absolute value | X
n-B
n| greater than the threshold value TH2 B of this sampled point simultaneously of regulation
nNearest non-of value and this sampled point exists the absolute value of gray-scale value of sampled point less than the threshold value TH3 of regulation, just be judged to be foreground object and entered on this sampled point, the renewal of this sampled point at this moment just with the nearest non-gray-scale value of sampled point that exists of this sampled point as B
N+1The background modeling of all the other sampled points all by formula (2) upgrades processing;
In Se and Sf calculation procedure, be used in each the threshold value TH1 that sets in the Sa step and carry out binary conversion treatment, all sampled points will be divided into " 0 " or " 1 " two states, have foreground object to exist on this sampled point of the expression of " 1 ", not have foreground object on this sampled point of the expression of " 0 ";
In the calculation procedure of Sg, adopt auto model to eliminate the misinterpretation of these sampled points; Specific algorithm is: exist sampled point around a non-situation that has sampled point is arranged, should non-ly exist sampled point according to auto model as misinterpretation, should non-ly exist sampled point to be modified to and have sampled point; Not existing around under the situation of sampled point has to have a sampled point; Exist the situation of foreground object on the described sampled point that exists sampled point to be illustrated in the parking stall, do not exist the situation of foreground object on the described non-sampled point that exists sampled point to be illustrated in the parking stall;
In the calculation procedure of Sh, all the sampling number purposes that exist in sampled point number and the parking stall in the calculation procedure according to above-mentioned Sg on resultant certain parking stall judge recently whether this parking stall occupied, computing method as shown in Equation (3),
Rate＝(T
ec/T
c)×100 (3)
In the formula, T
EcBe the statistical value that has sampled point in certain parking stall, T
cBe the statistical value of all sampled points in certain parking stall, Rate is the space hold ratio in certain parking stall;
If calculated value Rate, is judged to be this parking stall more than or equal to threshold value TH4 and is taken by vehicle.
5. the curb parking level detecting apparatus based on computer vision as claimed in claim 1, it is characterized in that: the employing circular arc minute surface of described panoramic vision sensor, described video camera is f to the distance of circular arc minute surface, the camera field of view scope is W * R, the panoramic vision sensor setting height(from bottom) is H, along the visual length on the road direction is L, the radius-of-curvature of circular arc minute surface is W, the central angle of circular arc is 60 °, the arc length of circular arc minute surface is π * W/3, ε be the circular arc minute surface on side view with the angle of surface level, η is the angle between video camera and the vertical rod.
6. as the described curb parking level detecting apparatus based on computer vision of one of claim 1～3, it is characterized in that: the naming method of described virtual parking stall: all parking stalls are with one group of array representation, PS
i* (n), n with in the face of the shooting direction of panoramic vision sensor from left to right with 1～10 series arrangement, PS
iBe illustrated on this road from left to right i panoramic vision sensor, * option have only two, N and A, wherein N represents the road side near panoramic vision sensor, A represents the road side away from panoramic vision sensor.
7. as the described curb parking level detecting apparatus of one of claim 1～3 based on computer vision, it is characterized in that: in the status information release module of described parking stall, all parking stall states in the parking place are sent to the management of parking charge system, and described the management of parking charge system chargeed according to the parking stall time that a certain car occupies.
8. as the described curb parking level detecting apparatus based on computer vision of one of claim 1～3, it is characterized in that: described parking guidance device is that road traffic is induced display board.
9. as the described curb parking level detecting apparatus of one of claim 1～3 based on computer vision, it is characterized in that: described parking guidance device is the webpage of intelligent transportation, and the mode that the user can browsing page is understood the situation that takies on parking stall on the road of somewhere.
10. as the described curb parking level detecting apparatus of one of claim 1～3 based on computer vision, it is characterized in that: described parking guidance device is the various service units of mobile Internet, and the user understands the situation that takies on parking stall on the road of somewhere in real time by short message mode.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2011100406740A CN102110376B (en) | 2011-02-18 | 2011-02-18 | Roadside parking space detection device based on computer vision |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2011100406740A CN102110376B (en) | 2011-02-18 | 2011-02-18 | Roadside parking space detection device based on computer vision |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102110376A true CN102110376A (en) | 2011-06-29 |
| CN102110376B CN102110376B (en) | 2012-11-21 |
Family
ID=44174518
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2011100406740A Expired - Fee Related CN102110376B (en) | 2011-02-18 | 2011-02-18 | Roadside parking space detection device based on computer vision |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102110376B (en) |
Cited By (49)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102663357A (en) * | 2012-03-28 | 2012-09-12 | 北京工业大学 | Color characteristic-based detection algorithm for stall at parking lot |
| CN103336268A (en) * | 2013-06-14 | 2013-10-02 | 北京航空航天大学 | Induction type non-contact charging position alignment device and method |
| CN103345266A (en) * | 2013-06-12 | 2013-10-09 | 西安应用光学研究所 | Vehicle-mounted photoelectricity visual guidance method based on panoramic picture |
| CN103366602A (en) * | 2012-03-29 | 2013-10-23 | 施乐公司 | Method of determining parking lot occupancy from digital camera images |
| CN103824452A (en) * | 2013-11-22 | 2014-05-28 | 银江股份有限公司 | Lightweight peccancy parking detection device based on full view vision |
| CN104376742A (en) * | 2014-12-02 | 2015-02-25 | 深圳市捷顺科技实业股份有限公司 | Parking lot state detection method and system |
| CN104834886A (en) * | 2014-02-11 | 2015-08-12 | 浙江大华技术股份有限公司 | Method and device for detecting video image |
| CN105678264A (en) * | 2016-01-07 | 2016-06-15 | 周丽娜 | Parking lot management system excellent in profile identification and filtering performances |
| CN105702087A (en) * | 2016-04-28 | 2016-06-22 | 武汉全华光电科技股份有限公司 | Intelligent parking system based on city street lamps and method thereof |
| CN105809184A (en) * | 2015-10-30 | 2016-07-27 | 哈尔滨工程大学 | Vehicle real-time identification tracking and parking space occupancy determining method suitable for gas station |
| CN105810007A (en) * | 2016-03-29 | 2016-07-27 | 北京小米移动软件有限公司 | Method and device for stopping balance car |
| CN105957397A (en) * | 2016-06-24 | 2016-09-21 | 齐鲁工业大学 | Intelligent parking integrated system based on Internet of things |
| CN105989629A (en) * | 2015-08-31 | 2016-10-05 | 湖南长宜物联科技有限公司 | Lane occupation parking time recording payment method |
| CN106023594A (en) * | 2016-06-13 | 2016-10-12 | 北京精英智通科技股份有限公司 | Parking stall shielding determination method and device and vehicle management system |
| CN106096554A (en) * | 2016-06-13 | 2016-11-09 | 北京精英智通科技股份有限公司 | Decision method and system are blocked in a kind of parking stall |
| CN106097722A (en) * | 2016-06-16 | 2016-11-09 | 广州地理研究所 | Video is utilized to carry out the system and method for trackside parking stall automatization supervision |
| CN106355886A (en) * | 2016-08-31 | 2017-01-25 | 智慧互通科技有限公司 | Parking management system for open type parking lot and management method of parking management system |
| CN106355869A (en) * | 2015-07-24 | 2017-01-25 | 徐工集团工程机械股份有限公司 | Vehicle scheduling method and system |
| CN106710299A (en) * | 2017-01-03 | 2017-05-24 | 山东浪潮商用系统有限公司 | Open pavement intelligent parking management method |
| CN106971602A (en) * | 2017-03-29 | 2017-07-21 | 深圳市金溢科技股份有限公司 | The condition detection method and laser car test equipment of a kind of parking position |
| CN106997685A (en) * | 2017-05-16 | 2017-08-01 | 刘程秀 | A kind of roadside parking space detection device based on microcomputerized visual |
| CN107134145A (en) * | 2017-06-10 | 2017-09-05 | 智慧互通科技有限公司 | Roadside Parking managing device, system and method based on polymorphic type IMAQ |
| CN107301788A (en) * | 2017-08-17 | 2017-10-27 | 京东方科技集团股份有限公司 | Virtual parking area parking method and system |
| CN107464450A (en) * | 2017-09-11 | 2017-12-12 | 浙江志诚软件有限公司 | A kind of road parking data collecting system |
| CN107610499A (en) * | 2016-07-11 | 2018-01-19 | 富士通株式会社 | Detection method, detection means and the electronic equipment of parking stall state |
| CN107749946A (en) * | 2017-09-28 | 2018-03-02 | 宁波优泊停车服务有限公司 | Road parking berth image pickup method, system, computer installation and computer-readable recording medium |
| CN108091168A (en) * | 2017-12-27 | 2018-05-29 | 深圳市道路交通管理事务中心 | A kind of parking management method and system |
| CN108091167A (en) * | 2017-12-27 | 2018-05-29 | 深圳市道路交通管理事务中心 | A kind of parking management method and system |
| CN108257408A (en) * | 2016-12-29 | 2018-07-06 | 杭州海存信息技术有限公司 | Collaborative parking position monitoring system |
| CN108257387A (en) * | 2017-12-27 | 2018-07-06 | 深圳市道路交通管理事务中心 | A kind of parking management method and system |
| CN108537894A (en) * | 2018-02-28 | 2018-09-14 | 深圳市元征科技股份有限公司 | A kind of parking charging method and device |
| CN106056968B (en) * | 2016-07-29 | 2018-09-21 | 北京华航无线电测量研究所 | A kind of method for detecting parking stalls based on optical imagery |
| WO2018177192A1 (en) * | 2017-03-31 | 2018-10-04 | 杭州海康威视数字技术股份有限公司 | State detection method for parallel parking and camera |
| CN108765973A (en) * | 2018-06-01 | 2018-11-06 | 智慧互通科技有限公司 | A kind of Roadside Parking management system based on the complementation of offside visual angle |
| CN108921956A (en) * | 2018-07-03 | 2018-11-30 | 重庆邮电大学 | A kind of curb parking charge management method based on Video Analysis Technology |
| CN109416882A (en) * | 2016-07-08 | 2019-03-01 | 罗伯特·博世有限公司 | In the determination of the separate parking space in side |
| CN109472184A (en) * | 2017-09-08 | 2019-03-15 | 深圳市金溢科技股份有限公司 | The condition detection method in berth, system and its data processing equipment in road |
| CN109817013A (en) * | 2018-12-19 | 2019-05-28 | 新大陆数字技术股份有限公司 | Parking stall state identification method and device based on video flowing |
| US10362112B2 (en) | 2014-03-06 | 2019-07-23 | Verizon Patent And Licensing Inc. | Application environment for lighting sensory networks |
| US10417570B2 (en) | 2014-03-06 | 2019-09-17 | Verizon Patent And Licensing Inc. | Systems and methods for probabilistic semantic sensing in a sensory network |
| CN110264580A (en) * | 2019-05-29 | 2019-09-20 | 深圳市轱辘汽车维修技术有限公司 | A kind of management method and device of on-fixed parking stall |
| CN110689761A (en) * | 2019-12-11 | 2020-01-14 | 上海赫千电子科技有限公司 | Automatic parking method |
| CN110706504A (en) * | 2019-08-27 | 2020-01-17 | 张雪华 | Intelligent parking space control method and system based on Internet of vehicles |
| CN111325858A (en) * | 2020-03-06 | 2020-06-23 | 赛特斯信息科技股份有限公司 | Method for realizing automatic charging management aiming at roadside temporary parking space |
| CN111488977A (en) * | 2019-01-25 | 2020-08-04 | 北京地平线机器人技术研发有限公司 | Neural network model training method and device |
| CN112016517A (en) * | 2020-09-14 | 2020-12-01 | 西安莱奥信息科技有限公司 | Parking space identification method and device based on machine vision |
| CN112634650A (en) * | 2020-12-18 | 2021-04-09 | 中标慧安信息技术股份有限公司 | Parking lot management method and system based on audio and video monitoring |
| CN113920770A (en) * | 2020-07-07 | 2022-01-11 | 北京新能源汽车股份有限公司 | Passenger-riding parking control method, device, equipment and vehicle |
| CN114040104A (en) * | 2021-11-16 | 2022-02-11 | 北京筑梦园科技有限公司 | Equipment debugging method and device and parking management system |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9374870B2 (en) | 2012-09-12 | 2016-06-21 | Sensity Systems Inc. | Networked lighting infrastructure for sensing applications |
| US9582671B2 (en) | 2014-03-06 | 2017-02-28 | Sensity Systems Inc. | Security and data privacy for lighting sensory networks |
| WO2014160708A1 (en) | 2013-03-26 | 2014-10-02 | Sensity Systems, Inc. | Sensor nodes with multicast transmissions in lighting sensory network |
| US9933297B2 (en) | 2013-03-26 | 2018-04-03 | Sensity Systems Inc. | System and method for planning and monitoring a light sensory network |
| US9746370B2 (en) | 2014-02-26 | 2017-08-29 | Sensity Systems Inc. | Method and apparatus for measuring illumination characteristics of a luminaire |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2002183897A (en) * | 2000-12-11 | 2002-06-28 | Taisei Corp | Parking state monitor system, parking situation sensing and display device in mobile parking lot |
| CN101059909A (en) * | 2006-04-21 | 2007-10-24 | 浙江工业大学 | All-round computer vision-based electronic parking guidance system |
| CN101064065A (en) * | 2007-03-29 | 2007-10-31 | 汤一平 | Parking inducing system based on computer visual sense |
-
2011
- 2011-02-18 CN CN2011100406740A patent/CN102110376B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2002183897A (en) * | 2000-12-11 | 2002-06-28 | Taisei Corp | Parking state monitor system, parking situation sensing and display device in mobile parking lot |
| CN101059909A (en) * | 2006-04-21 | 2007-10-24 | 浙江工业大学 | All-round computer vision-based electronic parking guidance system |
| CN101064065A (en) * | 2007-03-29 | 2007-10-31 | 汤一平 | Parking inducing system based on computer visual sense |
Cited By (68)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102663357A (en) * | 2012-03-28 | 2012-09-12 | 北京工业大学 | Color characteristic-based detection algorithm for stall at parking lot |
| CN103366602A (en) * | 2012-03-29 | 2013-10-23 | 施乐公司 | Method of determining parking lot occupancy from digital camera images |
| CN103345266A (en) * | 2013-06-12 | 2013-10-09 | 西安应用光学研究所 | Vehicle-mounted photoelectricity visual guidance method based on panoramic picture |
| CN103345266B (en) * | 2013-06-12 | 2015-09-23 | 西安应用光学研究所 | Based on the vehicular photoelectric visual guide method of panoramic picture |
| CN103336268B (en) * | 2013-06-14 | 2015-07-15 | 北京航空航天大学 | Induction type non-contact charging position alignment device and method |
| CN103336268A (en) * | 2013-06-14 | 2013-10-02 | 北京航空航天大学 | Induction type non-contact charging position alignment device and method |
| CN103824452B (en) * | 2013-11-22 | 2016-06-22 | 银江股份有限公司 | A kind of peccancy parking detector based on panoramic vision of lightweight |
| CN103824452A (en) * | 2013-11-22 | 2014-05-28 | 银江股份有限公司 | Lightweight peccancy parking detection device based on full view vision |
| CN104834886A (en) * | 2014-02-11 | 2015-08-12 | 浙江大华技术股份有限公司 | Method and device for detecting video image |
| US10791175B2 (en) | 2014-03-06 | 2020-09-29 | Verizon Patent And Licensing Inc. | Application environment for sensory networks |
| US10417570B2 (en) | 2014-03-06 | 2019-09-17 | Verizon Patent And Licensing Inc. | Systems and methods for probabilistic semantic sensing in a sensory network |
| US11616842B2 (en) | 2014-03-06 | 2023-03-28 | Verizon Patent And Licensing Inc. | Application environment for sensory networks |
| US10362112B2 (en) | 2014-03-06 | 2019-07-23 | Verizon Patent And Licensing Inc. | Application environment for lighting sensory networks |
| US11544608B2 (en) | 2014-03-06 | 2023-01-03 | Verizon Patent And Licensing Inc. | Systems and methods for probabilistic semantic sensing in a sensory network |
| CN104376742A (en) * | 2014-12-02 | 2015-02-25 | 深圳市捷顺科技实业股份有限公司 | Parking lot state detection method and system |
| CN106355869A (en) * | 2015-07-24 | 2017-01-25 | 徐工集团工程机械股份有限公司 | Vehicle scheduling method and system |
| CN105989629A (en) * | 2015-08-31 | 2016-10-05 | 湖南长宜物联科技有限公司 | Lane occupation parking time recording payment method |
| CN105809184A (en) * | 2015-10-30 | 2016-07-27 | 哈尔滨工程大学 | Vehicle real-time identification tracking and parking space occupancy determining method suitable for gas station |
| CN105809184B (en) * | 2015-10-30 | 2020-06-16 | 哈尔滨工程大学 | Method for real-time vehicle identification and tracking and parking space occupation judgment suitable for gas station |
| CN105678264A (en) * | 2016-01-07 | 2016-06-15 | 周丽娜 | Parking lot management system excellent in profile identification and filtering performances |
| CN105810007A (en) * | 2016-03-29 | 2016-07-27 | 北京小米移动软件有限公司 | Method and device for stopping balance car |
| WO2017166550A1 (en) * | 2016-03-29 | 2017-10-05 | 北京小米移动软件有限公司 | Balance car stopping method and device |
| US10336319B2 (en) | 2016-03-29 | 2019-07-02 | Beijing Xiaomi Mobile Software Co., Ltd. | Method, device and computer-readable storage medium for parking a self-balancing vehicle |
| CN105702087A (en) * | 2016-04-28 | 2016-06-22 | 武汉全华光电科技股份有限公司 | Intelligent parking system based on city street lamps and method thereof |
| CN106023594A (en) * | 2016-06-13 | 2016-10-12 | 北京精英智通科技股份有限公司 | Parking stall shielding determination method and device and vehicle management system |
| CN106096554A (en) * | 2016-06-13 | 2016-11-09 | 北京精英智通科技股份有限公司 | Decision method and system are blocked in a kind of parking stall |
| CN106097722B (en) * | 2016-06-16 | 2019-03-08 | 广州地理研究所 | The system and method for carrying out the automation supervision of trackside parking stall using video |
| CN106097722A (en) * | 2016-06-16 | 2016-11-09 | 广州地理研究所 | Video is utilized to carry out the system and method for trackside parking stall automatization supervision |
| CN105957397A (en) * | 2016-06-24 | 2016-09-21 | 齐鲁工业大学 | Intelligent parking integrated system based on Internet of things |
| CN109416882B (en) * | 2016-07-08 | 2022-06-14 | 罗伯特·博世有限公司 | Determination of laterally remote parking spaces |
| CN109416882A (en) * | 2016-07-08 | 2019-03-01 | 罗伯特·博世有限公司 | In the determination of the separate parking space in side |
| CN107610499A (en) * | 2016-07-11 | 2018-01-19 | 富士通株式会社 | Detection method, detection means and the electronic equipment of parking stall state |
| CN106056968B (en) * | 2016-07-29 | 2018-09-21 | 北京华航无线电测量研究所 | A kind of method for detecting parking stalls based on optical imagery |
| CN106355886A (en) * | 2016-08-31 | 2017-01-25 | 智慧互通科技有限公司 | Parking management system for open type parking lot and management method of parking management system |
| CN106355886B (en) * | 2016-08-31 | 2019-11-15 | 智慧互通科技有限公司 | A kind of open type parking ground parking management system and its management method |
| CN108257408A (en) * | 2016-12-29 | 2018-07-06 | 杭州海存信息技术有限公司 | Collaborative parking position monitoring system |
| CN106710299A (en) * | 2017-01-03 | 2017-05-24 | 山东浪潮商用系统有限公司 | Open pavement intelligent parking management method |
| CN106971602A (en) * | 2017-03-29 | 2017-07-21 | 深圳市金溢科技股份有限公司 | The condition detection method and laser car test equipment of a kind of parking position |
| WO2018177192A1 (en) * | 2017-03-31 | 2018-10-04 | 杭州海康威视数字技术股份有限公司 | State detection method for parallel parking and camera |
| CN106997685A (en) * | 2017-05-16 | 2017-08-01 | 刘程秀 | A kind of roadside parking space detection device based on microcomputerized visual |
| CN107134145A (en) * | 2017-06-10 | 2017-09-05 | 智慧互通科技有限公司 | Roadside Parking managing device, system and method based on polymorphic type IMAQ |
| CN107301788A (en) * | 2017-08-17 | 2017-10-27 | 京东方科技集团股份有限公司 | Virtual parking area parking method and system |
| CN109472184A (en) * | 2017-09-08 | 2019-03-15 | 深圳市金溢科技股份有限公司 | The condition detection method in berth, system and its data processing equipment in road |
| CN107464450A (en) * | 2017-09-11 | 2017-12-12 | 浙江志诚软件有限公司 | A kind of road parking data collecting system |
| CN107749946A (en) * | 2017-09-28 | 2018-03-02 | 宁波优泊停车服务有限公司 | Road parking berth image pickup method, system, computer installation and computer-readable recording medium |
| CN108257387A (en) * | 2017-12-27 | 2018-07-06 | 深圳市道路交通管理事务中心 | A kind of parking management method and system |
| CN108091168A (en) * | 2017-12-27 | 2018-05-29 | 深圳市道路交通管理事务中心 | A kind of parking management method and system |
| CN108091167A (en) * | 2017-12-27 | 2018-05-29 | 深圳市道路交通管理事务中心 | A kind of parking management method and system |
| CN108091168B (en) * | 2017-12-27 | 2020-07-31 | 深圳市道路交通管理事务中心 | Parking management method and system |
| CN108257387B (en) * | 2017-12-27 | 2020-07-31 | 深圳市道路交通管理事务中心 | Parking management method and system |
| CN108091167B (en) * | 2017-12-27 | 2020-07-03 | 深圳市道路交通管理事务中心 | Parking management method and system |
| CN108537894A (en) * | 2018-02-28 | 2018-09-14 | 深圳市元征科技股份有限公司 | A kind of parking charging method and device |
| CN108765973A (en) * | 2018-06-01 | 2018-11-06 | 智慧互通科技有限公司 | A kind of Roadside Parking management system based on the complementation of offside visual angle |
| CN108765973B (en) * | 2018-06-01 | 2022-04-19 | 智慧互通科技股份有限公司 | Roadside parking management system based on opposite side visual angle complementation |
| CN108921956A (en) * | 2018-07-03 | 2018-11-30 | 重庆邮电大学 | A kind of curb parking charge management method based on Video Analysis Technology |
| CN109817013A (en) * | 2018-12-19 | 2019-05-28 | 新大陆数字技术股份有限公司 | Parking stall state identification method and device based on video flowing |
| CN111488977A (en) * | 2019-01-25 | 2020-08-04 | 北京地平线机器人技术研发有限公司 | Neural network model training method and device |
| CN111488977B (en) * | 2019-01-25 | 2023-11-07 | 北京地平线机器人技术研发有限公司 | Neural network model training method and device |
| CN110264580A (en) * | 2019-05-29 | 2019-09-20 | 深圳市轱辘汽车维修技术有限公司 | A kind of management method and device of on-fixed parking stall |
| CN110264580B (en) * | 2019-05-29 | 2021-06-11 | 深圳市轱辘车联数据技术有限公司 | Management method and device for non-fixed parking spaces |
| CN110706504A (en) * | 2019-08-27 | 2020-01-17 | 张雪华 | Intelligent parking space control method and system based on Internet of vehicles |
| CN110706504B (en) * | 2019-08-27 | 2021-11-05 | 云南因特飞翔科技有限公司 | Intelligent parking space control method and system based on Internet of vehicles |
| CN110689761A (en) * | 2019-12-11 | 2020-01-14 | 上海赫千电子科技有限公司 | Automatic parking method |
| CN111325858A (en) * | 2020-03-06 | 2020-06-23 | 赛特斯信息科技股份有限公司 | Method for realizing automatic charging management aiming at roadside temporary parking space |
| CN113920770A (en) * | 2020-07-07 | 2022-01-11 | 北京新能源汽车股份有限公司 | Passenger-riding parking control method, device, equipment and vehicle |
| CN112016517A (en) * | 2020-09-14 | 2020-12-01 | 西安莱奥信息科技有限公司 | Parking space identification method and device based on machine vision |
| CN112634650A (en) * | 2020-12-18 | 2021-04-09 | 中标慧安信息技术股份有限公司 | Parking lot management method and system based on audio and video monitoring |
| CN114040104A (en) * | 2021-11-16 | 2022-02-11 | 北京筑梦园科技有限公司 | Equipment debugging method and device and parking management system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102110376B (en) | 2012-11-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102110376B (en) | Roadside parking space detection device based on computer vision | |
| KR102197946B1 (en) | object recognition and counting method using deep learning artificial intelligence technology | |
| CN101710448B (en) | Road traffic state detecting device based on omnibearing computer vision | |
| US20030123703A1 (en) | Method for monitoring a moving object and system regarding same | |
| US20030053659A1 (en) | Moving object assessment system and method | |
| US20030053658A1 (en) | Surveillance system and methods regarding same | |
| CN105611244A (en) | Method for detecting airport foreign object debris based on monitoring video of dome camera | |
| CN103605967A (en) | Subway fare evasion prevention system and working method thereof based on image recognition | |
| CN102819764A (en) | Method for counting pedestrian flow from multiple views under complex scene of traffic junction | |
| CN102136196A (en) | Vehicle velocity measurement method based on image characteristics | |
| Zheng et al. | Vehicle detection based on morphology from highway aerial images | |
| KR102122850B1 (en) | Solution for analysis road and recognition vehicle license plate employing deep-learning | |
| CN106339657A (en) | Straw incineration monitoring method and device based on monitoring video | |
| CN104504363A (en) | Real-time identification method of sidewalk on the basis of time-space correlation | |
| CN105262983A (en) | Road monitoring system and method based on internet of lamps | |
| CN107221175A (en) | A kind of pedestrian is intended to detection method and system | |
| CN104159088A (en) | System and method of remote monitoring of intelligent vehicle | |
| Ua-Areemitr et al. | Low-cost road traffic state estimation system using time-spatial image processing | |
| Shafie et al. | Smart video surveillance system for vehicle detection and traffic flow control | |
| CN109977796A (en) | Trail current detection method and device | |
| Chen et al. | A Novel Background Filtering Method with Automatic Parameter Adjustment for Real-Time Roadside LiDAR Sensing System | |
| Li et al. | Intelligent transportation video tracking technology based on computer and image processing technology | |
| KR102240638B1 (en) | Parking guidance method and system using boundary pixel data estimated in vehicle image and analysis of vehicle model viewpoint | |
| CN104463913A (en) | Intelligent illegal parking detection device and method | |
| Zhiwei et al. | Models of vehicle speeds measurement with a single camera |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C17 | Cessation of patent right | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20121121 Termination date: 20140218 |