CN109685000A

CN109685000A - A kind of method for detecting parking stalls and device of view-based access control model

Info

Publication number: CN109685000A
Application number: CN201811579916.1A
Authority: CN
Inventors: 张雪飞; 肖志光; 赖健明; 蒋少峰
Original assignee: Guangzhou Xiaopeng Motors Technology Co Ltd
Current assignee: Guangzhou Xiaopeng Motors Technology Co Ltd
Priority date: 2018-12-21
Filing date: 2018-12-21
Publication date: 2019-04-26
Also published as: WO2020124988A1

Abstract

The present embodiments relate to parking stall measure technical fields, disclose the method for detecting parking stalls and device of a kind of view-based access control model, this method comprises: being pre-processed to the ambient image of the vehicle two sides taken, to obtain corresponding aerial view, it detects the parking stall frame and parking stall angle point in aerial view and screens a target parking stall frame from the parking stall frame in the aerial view, according to the target parking stall frame, from the target carriage parallactic angle point obtained in the angle point of parking stall on the target parking stall frame, according to the position coordinates of the target carriage parallactic angle point in aerial view, determine the corresponding true location coordinate of target carriage parallactic angle point, according to the corresponding true location coordinate of target carriage parallactic angle point, determine the corresponding target parking stall of target parking stall frame.It can be improved the accuracy of parking stall measure, to improve the discrimination of parking stall.

Description

A kind of method for detecting parking stalls and device of view-based access control model

Technical field

The present invention relates to parking stall measure technical field more particularly to the method for detecting parking stalls and device of a kind of view-based access control model.

Background technique

In recent years, with the rapid development of economy, the quantity of automobile increases sharply, but the limited amount on parking stall, this gives It finds parking stall and brings certain difficulty, the development of parking stall measure technology solves find this hardly possible of parking stall to a certain extent Topic.Currently, the mode of parking stall measure has following several method: the method for the cutting techniques identification parking stall based on color is based on Canny edge detection and Hough straight-line detection technology are identified the method for parking stall and extracted based on SIFT feature to be known with matching technique The method of other parking stall.But the resolution of the above-mentioned method that parking stall is identified based on parking stall line color or shape is lower, identifies The success rate of parking stall is not high.Therefore, it needs to develop a kind of higher method for detecting parking stalls of parking stall discrimination.

Summary of the invention

The embodiment of the invention discloses a kind of method for detecting parking stalls of view-based access control model and devices, can be improved parking stall measure Accuracy, to improve the discrimination of parking stall.

First aspect of the embodiment of the present invention discloses a kind of method for detecting parking stalls of view-based access control model, comprising:

The ambient image of the vehicle two sides taken is pre-processed, to obtain corresponding aerial view；

Detect the parking stall frame and parking stall angle point in the aerial view；

A target parking stall frame is screened from the parking stall frame in the aerial view；

According to target parking stall frame, from the target carriage parallactic angle obtained in the parking stall angle point on the frame of the target parking stall Point；

According to the position coordinates of the target carriage parallactic angle point in the aerial view, the target carriage parallactic angle point pair is determined The true location coordinate answered；

According to the corresponding true location coordinate of the target carriage parallactic angle point, the corresponding target carriage of target parking stall frame is determined Position.

As an alternative embodiment, in first aspect of the embodiment of the present invention, the described pair of vehicle two taken The ambient image of side is pre-processed, to obtain corresponding aerial view, comprising:

According to the inner parameter and external parameter of the photographic device of vehicle, to the environment map of the vehicle two sides taken As carrying out distortion correction, to obtain corresponding correcting image；

According to the pixel of the correcting image, perspective transform is carried out to the correcting image, to obtain corresponding get a bird's eye view View.

As an alternative embodiment, in first aspect of the embodiment of the present invention, in the camera shooting according to vehicle The inner parameter and external parameter of device carry out distortion correction to the ambient image of the vehicle two sides taken, to obtain Before corresponding correcting image, the method also includes:

Obtain the inner parameter and external parameter of the photographic device of vehicle；The inner parameter is filled including at least the camera shooting Focal length, Pixel Dimensions and the distortion parameter set；The external parameter includes at least height of the photographic device relative to ground With the spin matrix of the photographic device；

The inner parameter and external parameter of the photographic device according to vehicle, to the ring of the vehicle two sides taken Border image carries out distortion correction, to obtain corresponding correcting image, comprising:

According to the inner parameter and the external parameter, the ambient image and corresponding correction of the vehicle two sides are determined The first corresponding relationship between image；First corresponding relationship includes the picture of the ambient image of the vehicle two sides taken Corresponding relationship between vegetarian refreshments and the pixel of the correcting image；

First corresponding relationship is saved to the first look-up table；

According to first look-up table, distortion correction is carried out to the ambient image of the vehicle two sides taken, to obtain Obtain the correcting image；

The pixel according to the correcting image carries out perspective transform to the correcting image, corresponding to obtain Aerial view, comprising:

According to the pixel of the correcting image, second pair between the correcting image and corresponding aerial view is determined It should be related to；Second corresponding relationship includes pair between the pixel of the correcting image and the pixel of the aerial view It should be related to；

Second corresponding relationship is saved to second look-up table；

According to the second look-up table, perspective transform is carried out to the correcting image, to obtain the aerial view.

As an alternative embodiment, in first aspect of the embodiment of the present invention, in the vehicle that described pair takes The ambient image of two sides is pre-processed, before obtaining corresponding aerial view, the method also includes:

And the ambient image of the described pair of vehicle two sides taken pre-processes, to obtain corresponding aerial view, Include:

According to the inner parameter and the external parameter, determines the ambient image of the vehicle two sides and corresponding get a bird's eye view Third corresponding relationship between view；The third corresponding relationship includes the picture of the ambient image of the vehicle two sides taken Corresponding relationship between vegetarian refreshments and the pixel of the aerial view；

The third corresponding relationship is saved to third look-up table；

According to the third look-up table, the ambient image of the vehicle two sides taken is carried out getting a bird's eye view transformation, to obtain Obtain the aerial view.

As an alternative embodiment, in first aspect of the embodiment of the present invention, described according to the target carriage Position frame, it is after the target carriage parallactic angle point obtained in the parking stall angle point on the frame of the target parking stall and described according to The position coordinates of the target carriage parallactic angle point in aerial view determine the corresponding true location coordinate of the target carriage parallactic angle point Before, the method also includes:

Whether there are obstacles in the region for detecting in target parking stall frame institute frame；

If there are barriers for the region in target parking stall frame institute frame, the area in target parking stall frame institute frame is marked Domain is can not parking stall；

If barrier is not present in the region in target parking stall frame institute frame, execute described according in the aerial view The target carriage parallactic angle point position coordinates, the step of determining the target carriage parallactic angle point corresponding true location coordinate.

Second aspect of the embodiment of the present invention discloses a kind of stall testing apparatus, comprising:

Pretreatment unit is pre-processed for the ambient image to the vehicle two sides taken, to obtain corresponding bird It looks down from a height view；

First detection unit, for detecting parking stall frame and parking stall angle point in the aerial view；

Screening unit, for screening a target parking stall frame from the parking stall frame in the aerial view；

Acquiring unit, for being obtained on the frame of the target parking stall from the parking stall angle point according to target parking stall frame Target carriage parallactic angle point；

First determination unit is determined for the position coordinates according to the target carriage parallactic angle point in the aerial view The corresponding true location coordinate of the target carriage parallactic angle point；

Second determination unit, for determining the target according to the corresponding true location coordinate of the target carriage parallactic angle point The corresponding target parking stall of parking stall frame.

As an alternative embodiment, in second aspect of the embodiment of the present invention, the pretreatment unit includes:

Distortion correction subelement, for the inner parameter and external parameter according to the photographic device of vehicle, to what is taken The ambient image of the vehicle two sides carries out distortion correction, to obtain corresponding correcting image；

Perspective transform subelement carries out perspective change to the correcting image for the pixel according to the correcting image It changes, to obtain corresponding aerial view.

As an alternative embodiment, the stall testing apparatus also wraps in second aspect of the embodiment of the present invention It includes:

First parameter acquiring unit, for the inner parameter in the distortion correction subelement according to the photographic device of vehicle And external parameter, distortion correction is carried out to the ambient image of the vehicle two sides taken, to obtain corresponding correcting image Before, the inner parameter and external parameter of the photographic device of vehicle are obtained；The inner parameter includes at least the photographic device Focal length, Pixel Dimensions and distortion parameter；The external parameter include at least the photographic device relative to ground height and The spin matrix of the photographic device；

The distortion correction subelement includes:

First determining module, for determining the ring of the vehicle two sides according to the inner parameter and the external parameter The first corresponding relationship between border image and corresponding correcting image；First corresponding relationship includes the vehicle taken Corresponding relationship between the pixel of the ambient image of two sides and the pixel of the correcting image；

First preserving module, for saving first corresponding relationship to the first look-up table；

Distortion correction module is used for according to first look-up table, to the ambient image of the vehicle two sides taken Distortion correction is carried out, to obtain the correcting image；

The perspective transform subelement includes:

Second determining module determines the correcting image and corresponding bird for the pixel according to the correcting image The second corresponding relationship looked down from a height between view；Second corresponding relationship includes the pixel of the correcting image and described gets a bird's eye view view Corresponding relationship between the pixel of figure；

Second preserving module is used for the second corresponding relationship second look-up table；

Perspective transform module, for carrying out perspective transform to the correcting image according to the second look-up table, to obtain The aerial view.

Second parameter acquiring unit, for being carried out in ambient image of the pretreatment unit to the vehicle two sides taken Pretreatment, before obtaining corresponding aerial view, obtains the inner parameter and external parameter of the photographic device of vehicle；In described Portion's parameter includes at least focal length, Pixel Dimensions and the distortion parameter of the photographic device；The external parameter includes at least described Photographic device is relative to the height on ground and the spin matrix of the photographic device；

The pretreatment unit includes:

Subelement is determined, for determining the environment of the vehicle two sides according to the inner parameter and the external parameter Third corresponding relationship between image and corresponding aerial view；The third corresponding relationship includes the vehicle two taken Corresponding relationship between the pixel of the ambient image of side and the pixel of the aerial view；

Saving subunit, for saving the third corresponding relationship to third look-up table；

Transformation subelement is got a bird's eye view, is used for according to the third look-up table, to the environment map of the vehicle two sides taken As carrying out getting a bird's eye view transformation, to obtain the aerial view.

Second detection unit, for, according to target parking stall frame, being obtained from the parking stall angle point in the acquiring unit It takes after the target carriage parallactic angle point on the frame of the target parking stall and first determination unit is according in the aerial view The position coordinates of the target carriage parallactic angle point before determining the corresponding true location coordinate of the target carriage parallactic angle point, detect institute Whether there are obstacles in the region for stating in target parking stall frame institute frame；

Marking unit, for detecting that the region in target parking stall frame institute frame has barrier in the second detection unit When hindering object, marking the region in target parking stall frame institute frame to be can not parking stall；

First determination unit, specifically for being detected in target parking stall frame institute frame in the second detection unit Region when barrier is not present, according to the position coordinates of the target carriage parallactic angle point in the aerial view, determine described in The corresponding true location coordinate of target carriage parallactic angle point.

The third aspect of the embodiment of the present invention discloses a kind of stall testing apparatus, comprising:

It is stored with the memory of executable program code；

The processor coupled with the memory；

The processor calls the executable program code stored in the memory, executes the embodiment of the present invention the A kind of method for detecting parking stalls of view-based access control model disclosed in one side.

Fourth aspect of the embodiment of the present invention discloses a kind of computer readable storage medium, stores computer program, wherein The computer program makes computer execute a kind of parking stall measure of view-based access control model disclosed in first aspect of the embodiment of the present invention Method.

The 5th aspect of the embodiment of the present invention discloses a kind of computer program product, when the computer program product is calculating When being run on machine, so that the computer executes some or all of any one method of first aspect step.

The aspect of the embodiment of the present invention the 6th disclose a kind of using distribution platform, and the application distribution platform is for publication calculating Machine program product, wherein when the computer program product is run on computers, so that the computer executes first party Some or all of any one method in face step.

Compared with prior art, the embodiment of the present invention has the advantages that

In the embodiment of the present invention, the ambient image of the vehicle two sides taken is pre-processed, to obtain corresponding bird It looks down from a height view, detect the parking stall frame and parking stall angle point in aerial view and screens a target from the parking stall frame in the aerial view Parking stall frame, according to the target parking stall frame, from the target carriage parallactic angle point obtained in the angle point of parking stall on the target parking stall frame, according to getting a bird's eye view The position coordinates of target carriage parallactic angle point in view determine the corresponding true location coordinate of target carriage parallactic angle point, according to target carriage The corresponding true location coordinate of parallactic angle point determines the corresponding target parking stall of target parking stall frame.As it can be seen that implement the embodiment of the present invention, It can be improved the accuracy of parking stall measure, to improve the discrimination of parking stall.

Detailed description of the invention

It to describe the technical solutions in the embodiments of the present invention more clearly, below will be to needed in the embodiment Attached drawing is briefly described, it should be apparent that, drawings in the following description are only some embodiments of the invention, for ability For the those of ordinary skill of domain, without creative efforts, it can also be obtained according to these attached drawings other attached Figure.

Fig. 1 is a kind of flow diagram of the method for detecting parking stalls of view-based access control model disclosed by the embodiments of the present invention；

Fig. 2 is the flow diagram of the method for detecting parking stalls of another view-based access control model disclosed by the embodiments of the present invention；

Fig. 3 is the flow diagram of the method for detecting parking stalls of another view-based access control model disclosed by the embodiments of the present invention；

Fig. 4 is a kind of structural schematic diagram of stall testing apparatus disclosed by the embodiments of the present invention；

Fig. 5 is the structural schematic diagram of another stall testing apparatus disclosed by the embodiments of the present invention；

Fig. 6 is the structural schematic diagram of another stall testing apparatus disclosed by the embodiments of the present invention；

Fig. 7 is the structural schematic diagram of another stall testing apparatus disclosed by the embodiments of the present invention；

Fig. 8 is a kind of application exemplary diagram of parking detection process disclosed by the embodiments of the present invention.

Specific embodiment

Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete Site preparation description, it is clear that the described embodiment is only a part of the embodiment of the present invention, instead of all the embodiments.Based on this Embodiment in invention, every other reality obtained by those of ordinary skill in the art without making creative efforts Example is applied, shall fall within the protection scope of the present invention.

It should be noted that term " first ", " second ", " third " in description and claims of this specification and " 4th " etc. is not use to describe a particular order for distinguishing different objects.The term " includes " of the embodiment of the present invention " having " and their any deformation, it is intended that cover it is non-exclusive include, for example, containing series of steps or list The process, method, system, product or equipment of member those of are not necessarily limited to be clearly listed step or unit, but may include not having There are other step or units being clearly listed or intrinsic for these process, methods, product or equipment.

The embodiment of the invention discloses a kind of method for detecting parking stalls of view-based access control model and devices, can be improved parking stall measure Accuracy, to improve the discrimination of parking stall.It is described in detail below in conjunction with attached drawing.

Embodiment one

Referring to Fig. 1, Fig. 1 is a kind of process signal of the method for detecting parking stalls of view-based access control model disclosed by the embodiments of the present invention Figure.As shown in Figure 1, this method may comprise steps of.

101, stall testing apparatus pre-processes the ambient image of the vehicle two sides taken, to obtain corresponding bird It looks down from a height view.

In the embodiment of the present invention, stall testing apparatus is by being arranged on the B column of vehicle two sides photographic device to vehicle two The environment of side is shot, to obtain the ambient image of vehicle two sides, also referring to Fig. 8, as shown in figure 8, Fig. 8 is the present invention A kind of application exemplary diagram of parking detection process disclosed in embodiment, vehicle are opened and are arranged in vehicle when carrying out parking stall measure Photographic device on the B column of two sides, photographic device shoot the environment of vehicle two sides, to obtain the environment of vehicle two sides Image.

102, the parking stall frame and parking stall angle point in stall testing apparatus detection aerial view.

In the embodiment of the present invention, stall testing apparatus can use the parking stall frame in deep learning algorithm detection aerial view With parking stall angle point, the embodiment of the present invention is not construed as limiting.

As an alternative embodiment, stall testing apparatus can acquire a large amount of parking stall figure sample, according to vehicle Bit pattern sample training obtains parking stall model, and above-mentioned aerial view is then imported the parking stall model and carries out convolution algorithm to generate Various features detect from various features according to default feature and obtain parking stall frame and parking stall angle point.Implement above embodiment, benefit The parking stall frame and parking stall angle point in aerial view are detected with deep learning algorithm, can be improved the precision of detection.

Further, as an alternative embodiment, stall testing apparatus can classify to feature, such as will Parking stall angle point is divided into 0 class, and parking stall frame is divided into 1 class, and above-mentioned aerial view is then imported the parking stall model and carries out convolution algorithm To generate various features, 0 class therein is extracted as parking stall angle point and extracts 1 class therein as parking stall Frame.Implement above embodiment, parking stall angle point is divided into 0 class and parking stall frame is divided into 1 class, can be improved the identification of computer Speed, to improve the detection efficiency of parking stall.

103, stall testing apparatus screens a target parking stall frame from the parking stall frame in aerial view.

As an alternative embodiment, step 103 stall testing apparatus is screened from the parking stall frame in aerial view One target parking stall frame, comprising:

The width of each parking stall frame in aerial view is ranked up and counts each parking stall frame by stall testing apparatus The ranking of width；

Stall testing apparatus will rank the first corresponding parking stall frame as target parking stall frame.

Implement above embodiment, filter out the maximum parking stall frame of width as target parking stall frame, can be improved and park Safety.

104, stall testing apparatus is according to target parking stall frame, from the target carriage obtained on the target parking stall frame in the angle point of parking stall Parallactic angle point.

In the embodiment of the present invention, stall testing apparatus can use traditional image processing method from the vehicle in aerial view A target parking stall frame is screened in the frame of position, and rejects other parking stall angles other than the target carriage parallactic angle point on the target parking stall frame Point.

In the embodiment of the present invention, the quantity at least two of the target carriage parallactic angle point on the frame of above-mentioned target parking stall, the present invention Embodiment is not construed as limiting.

105, stall testing apparatus determines the target parking stall according to the position coordinates of the target carriage parallactic angle point in aerial view The corresponding true location coordinate of angle point.

In the embodiment of the present invention, formula (1) expresses reflecting between the image coordinate system of aerial view and world coordinate system Relationship is penetrated, the position coordinates of the target carriage parallactic angle point in aerial view can be converted by stall testing apparatus according to formula (1) The position coordinates of target carriage parallactic angle point in world coordinate system, and using the position coordinates of target carriage parallactic angle point in world coordinate system as The corresponding true location coordinate of target carriage parallactic angle point, formula (1) are as follows:

Wherein, (X_w, Y_w, Z_w) be world coordinate system coordinate, (x, y) be aerial view image coordinate system coordinate, Z_c Height for photographic device relative to ground, f are the focal length of photographic device, and R is the spin matrix of photographic device.

106, stall testing apparatus determines that target parking stall frame is corresponding according to the corresponding true location coordinate of target carriage parallactic angle point Target parking stall.

In the embodiment of the present invention, stall testing apparatus determines mesh according to the corresponding true location coordinate of target carriage parallactic angle point The center in the region that mark parking stall angle point is constituted, the center in the region that target carriage parallactic angle point is constituted is as target The center of parking stall, and target carriage parallactic angle point is placed in the corner of target parking stall.

As it can be seen that implementing method described in Fig. 1, the ambient image of vehicle two sides is obtained by shooting, then by the environment Image is converted into aerial view, screens on a target parking stall frame and the target parking stall frame from the parking stall frame in aerial view Target carriage parallactic angle point, pass through to calculate and obtain the corresponding true location coordinate of target carriage parallactic angle point and determine that target parking stall frame is corresponding Target parking stall, can be improved the accuracy of parking stall measure, to improve the discrimination of parking stall.

Embodiment two

Referring to Fig. 2, the process that Fig. 2 is the method for detecting parking stalls of another view-based access control model disclosed by the embodiments of the present invention is shown It is intended to.As shown in Fig. 2, this method may comprise steps of.

201, stall testing apparatus obtains the inner parameter and external parameter of the photographic device of vehicle.

In the embodiment of the present invention, above-mentioned inner parameter includes at least focal length, Pixel Dimensions and the distortion parameter of photographic device, Said external parameter includes at least photographic device relative to the height on ground and the spin matrix of photographic device, the embodiment of the present invention It is not construed as limiting.

202, stall testing apparatus is according to the inner parameter and external parameter of the photographic device of vehicle, to the vehicle taken The ambient image of two sides carries out distortion correction, to obtain corresponding correcting image.

In the embodiment of the present invention, because pattern distortion will affect the detection of parking stall frame and parking stall angle point, to reduce vehicle Position detection accuracy, therefore need first to carry out the distortion correction of image before detection parking stall frame and parking stall angle point.The present invention is implemented In example, stall testing apparatus is according to the inner parameter and external parameter of the photographic device of vehicle, to the vehicle two sides taken It can be Zhang Zhengyou gridiron pattern standardization that ambient image, which carries out the method that distortion correction is taken, and the embodiment of the present invention does not limit It is fixed.

As an alternative embodiment, step 202 stall testing apparatus is joined according to the inside of the photographic device of vehicle Several and external parameter carries out distortion correction to the ambient image of the vehicle two sides taken, to obtain corresponding correcting image, packet It includes:

Stall testing apparatus determines the ambient image and corresponding correction of vehicle two sides according to inner parameter and external parameter The first corresponding relationship between image；

Stall testing apparatus saves the first corresponding relationship to the first look-up table；

Stall testing apparatus carries out distortion correction according to the first look-up table, to the ambient image of the vehicle two sides taken, To obtain corresponding correcting image.

In the embodiment of the present invention, the first corresponding relationship includes the pixel of the ambient image of the vehicle two sides taken and rectifys Corresponding relationship between the pixel of positive image.

In the embodiment of the present invention, the image that photographic device is shot can generate a degree of distortion, be filled according to camera shooting The inner parameter and external parameter set can calculate the corresponding relationship for correcting this distortion degree, that is to say, that parking stall measure Focal length f, the Pixel Dimensions (d of device acquisition photographic device_x×d_y) and distortion parameter k=[k₁, k₂], and, photographic device is opposite Height Z in ground_cWith the spin matrix R of photographic device, above-mentioned corresponding relationship is then calculated, wherein above-mentioned corresponding pass System can be indicated with formula (2):

Wherein, certain point coordinate in the world coordinate system of the environment of vehicle two sides is (X_w, Y_w, Z_w), the ring in vehicle two sides Coordinate in the image coordinate system in border is (x, y), the coordinate in the pixel coordinate system of the ambient image of vehicle two sides be (u, V), the coordinate in the pixel coordinate system of correcting image is (u ', v '), point (u₀, v₀) it is photographic device optical axis and the plane of delineation Intersection point, i.e. principal point, Z_cHeight for photographic device relative to ground, d_xFor the physical size of pixel in the direction of the x axis, d_yFor picture The physical size of element in the y-axis direction, k₁、k₂Preceding two rank of respectively distortion parameter k, f are the focal length of photographic device, and R is camera shooting The spin matrix of device.

In the embodiment of the present invention, stall testing apparatus can be according to formula (1), by the vehicle two sides in world coordinate system Coordinate points (X in environment_w, Y_w, Z_w) be converted to the vehicle two sides in pixel coordinate system ambient image pixel coordinate point (u, V), the corresponding pixel points (u ', v ') in correcting image then by each of image pixel coordinate point (u, v) are converted to, and will Corresponding pixel points (u ', v ') in each of ambient image of vehicle two sides pixel (u, v) and correcting image are stored in It is subsequent when needing to carry out distortion correction in first look-up table, the environment of vehicle two sides can be directly found according to the first look-up table Corresponding pixel points (u ', v ') of each of image pixel (u, v) in correcting image, then to the environment of vehicle two sides Image carries out distortion correction, to obtain correcting image.

203, stall testing apparatus carries out perspective transform to correcting image, with acquisition pair according to the pixel of correcting image The aerial view answered.

In the embodiment of the present invention, aerial view is the view that the corresponding actual environment of image is seen from top to bottom, because Be more advantageous to positioning of the stall testing apparatus to parking stall frame and parking stall angle point with aerial view presentation, thus in detection parking stall frame and It needs correcting image being converted to aerial view before the angle point of parking stall.

As an alternative embodiment, pixel of step 203 stall testing apparatus according to correcting image, to correction Image carries out perspective transform, to obtain corresponding aerial view, comprising:

Stall testing apparatus determines between correcting image and corresponding aerial view according to the pixel of correcting image Two corresponding relationships；

Stall testing apparatus saves the second corresponding relationship to second look-up table；

Stall testing apparatus carries out perspective transform according to second look-up table, to correcting image, corresponding gets a bird's eye view view to obtain Figure.

In the embodiment of the present invention, the second corresponding relationship includes between the pixel of correcting image and the pixel of aerial view Corresponding relationship.

In the embodiment of the present invention, stall testing apparatus can according to formula (3) by correcting image in pixel coordinate system Two-dimensional coordinate (u ', v ') is mapped to the three-dimensional coordinate (X in world coordinate system_w, Y_w, Z_w), then divided by Z in three-dimensional coordinate_wValue with Two-dimensional space between mapping back obtains coordinate (u ", v "), to obtain aerial view, formula (3) is as follows:

Wherein, (u ', v ') is the coordinate of the pixel coordinate system of correcting image, (X_w, Y_w, Z_w) it is that point (u ', v ') is sat in the world Three-dimensional coordinate in mark system, (u ", v ") are the coordinate of the pixel coordinate system of aerial view,For Transformation matrix, whereinFor linearity transformation, [a₁₃ a₂₃]^TFor perspective image transformation, [a₃₁ a₃₂] be Image translation in the embodiment of the present invention, gives transformation matrix, perspective transform can be carried out to correcting image, to be got a bird's eye view View.

In the embodiment of the present invention, stall testing apparatus can be according to formula (2), will be in the correcting image in pixel coordinate system Pixel coordinate point (u ', v ') be converted to the pixel coordinate point (u ", v ") of the aerial view in pixel coordinate system, and will correction figure Corresponding pixel points (u ", v ") in each of picture pixel (u ', v ') and aerial view are stored in second look-up table, after It is continuous when needing to carry out perspective transform to correcting image, each of correcting image picture directly can be found according to second look-up table Then corresponding pixel points (u ", v ") of the vegetarian refreshments (u ', v ') in aerial view carry out perspective transform to correcting image, to obtain Obtain aerial view.

204, the parking stall frame and parking stall angle point in stall testing apparatus detection aerial view.

205, stall testing apparatus screens a target parking stall frame from the parking stall frame in aerial view.

206, stall testing apparatus is according to target parking stall frame, from the target carriage obtained on the target parking stall frame in the angle point of parking stall Parallactic angle point.

207, whether there are obstacles for the region in stall testing apparatus detection target parking stall frame institute frame；If so, executing Step 208；If not, executing step 209.

As an alternative embodiment, stall testing apparatus can choose multiple regions at the edge of target parking stall frame And the marginal density of this multiple regions is detected, the marginal density in the maximum corresponding region of marginal density is close as object edge Degree, then judges whether object edge density is greater than specified threshold, if so, showing that there are barrier, parking stall measures in the region Region in device execution step 208 label target parking stall frame institute frame is can not parking stall；If not, stall testing apparatus executes Step 209 determines the corresponding true position of target carriage parallactic angle point according to the position coordinates of the target carriage parallactic angle point in aerial view Set coordinate.Implement above embodiment, it can be determined that in the region in target parking stall frame institute frame that stall testing apparatus filters out Whether there are obstacles, avoids colliding barrier during parking, can be improved safety of parking.

208, stall testing apparatus label target parking stall frame institute frame in region be can not parking stall.

In the embodiment of the present invention, if there are barriers in region in target parking stall frame institute frame, show target parking stall frame Corresponding parking stall cannot accommodate vehicle and moor into it is that can not stop that stall testing apparatus, which marks the region in the target parking stall institute frame, Position.

209, stall testing apparatus determines the target parking stall according to the position coordinates of the target carriage parallactic angle point in aerial view The corresponding true location coordinate of angle point.

210, stall testing apparatus determines that target parking stall frame is corresponding according to the corresponding true location coordinate of target carriage parallactic angle point Target parking stall.

As it can be seen that implement method described in Fig. 2 compared with the method described in Fig. 1, after determining target carriage parallactic angle point, Stall testing apparatus detects the region in target parking stall frame institute frame, and whether there are obstacles, if so, show that the parking stall can not stop, Stall testing apparatus mark the target parking stall frame enclose in region be can not parking stall, can be avoided and hit during vehicle parking To barrier, guarantee safety of parking.In addition, implementing method described in Fig. 2, by the first look-up table pair for carrying distortion parameter The ambient images of the vehicle two sides taken carries out distortion correction and obtains correcting image, then the by carrying perspective transform coefficient Two look-up tables carry out perspective transform to correcting image and obtain aerial view, can be improved the formation speed of aerial view, improve vehicle Position detection efficiency.

Embodiment three

Referring to Fig. 3, the process that Fig. 3 is the method for detecting parking stalls of another view-based access control model disclosed by the embodiments of the present invention is shown It is intended to.As shown in figure 3, this method may comprise steps of.

301, stall testing apparatus obtains the inner parameter and external parameter of the photographic device of vehicle.

302, stall testing apparatus is according to inner parameter and external parameter, determines the ambient image of vehicle two sides and corresponding Third corresponding relationship between aerial view.

In the embodiment of the present invention, third corresponding relationship includes pixel and the bird of the ambient image of the vehicle two sides taken The corresponding relationship looked down from a height between the pixel of view.

303, stall testing apparatus saves third corresponding relationship to third look-up table.

304, stall testing apparatus carries out getting a bird's eye view change according to third look-up table to the ambient image of the vehicle two sides taken It changes, to obtain corresponding aerial view.

In the embodiment of the present invention, stall testing apparatus obtains focal length f, the Pixel Dimensions (d of photographic device_x×d_y) and distortion Parameter k=[k₁, k₂], and, height Z of the photographic device relative to ground_cWith the spin matrix R of photographic device, then can tie The formula (2) stated and formula (3) is closed to turn the pixel coordinate point (u, v) of the ambient image of the vehicle two sides in pixel coordinate system It is melted into the pixel coordinate point (u ", v ") of the aerial view in pixel coordinate system, and will be each in the ambient image of vehicle two sides Corresponding pixel points (u ", v ") in a pixel (u, v) and aerial view are stored in third look-up table, subsequent to need to vehicle When the ambient image of two sides carries out getting a bird's eye view transformation, in ambient image that vehicle two sides can be directly found according to third look-up table Then corresponding pixel points (u ", v ") of each pixel (u, v) in aerial view carry out the ambient image of vehicle two sides Transformation is got a bird's eye view, to obtain aerial view.

305, the parking stall frame and parking stall angle point in stall testing apparatus detection aerial view.

306, stall testing apparatus screens a target parking stall frame from the parking stall frame in aerial view.

307, stall testing apparatus is according to target parking stall frame, from the target carriage obtained on the target parking stall frame in the angle point of parking stall Parallactic angle point.

308, whether there are obstacles for the region in stall testing apparatus detection target parking stall frame institute frame；If so, executing Step 309；If not, executing step 310.

309, stall testing apparatus label target parking stall frame institute frame in region be can not parking stall.

310, stall testing apparatus determines the target parking stall according to the position coordinates of the target carriage parallactic angle point in aerial view The corresponding true location coordinate of angle point.

311, stall testing apparatus determines that target parking stall frame is corresponding according to the corresponding true location coordinate of target carriage parallactic angle point Target parking stall.

As it can be seen that compared with the method described in Fig. 1, method described in implementing Fig. 3, after determining target carriage parallactic angle point, Stall testing apparatus detects the region in target parking stall frame institute frame, and whether there are obstacles, if so, show that the parking stall can not stop, Stall testing apparatus mark the target parking stall frame enclose in region be can not parking stall, can be avoided and hit during vehicle parking To barrier, guarantee safety of parking.In addition, method described in implementing Fig. 3, the third lookup of transformation coefficient is got a bird's eye view by carrying Table carries out the ambient image of the vehicle two sides taken to get a bird's eye view transformation acquisition aerial view, can be improved the generation of aerial view Speed improves parking stall measure efficiency.

Example IV

Referring to Fig. 4, Fig. 4 is a kind of structural schematic diagram of stall testing apparatus disclosed by the embodiments of the present invention.Such as Fig. 4 institute Show, which may include:

Pretreatment unit 401 is pre-processed for the ambient image to the vehicle two sides taken, corresponding to obtain Aerial view；

First detection unit 402, for detecting parking stall frame and parking stall angle point in aerial view；

Screening unit 403, for screening a target parking stall frame from the parking stall frame in aerial view；

Acquiring unit 404 is used for according to target parking stall frame, from the target carriage obtained on the target parking stall frame in the angle point of parking stall Parallactic angle point；

First determination unit 405 determines the target for the position coordinates according to the target carriage parallactic angle point in aerial view The corresponding true location coordinate of parking stall angle point；

Second determination unit 406, for determining target parking stall frame according to the corresponding true location coordinate of target carriage parallactic angle point Corresponding target parking stall.

In the embodiment of the present invention, first detection unit 402 can use the vehicle in deep learning algorithm detection aerial view Position frame and parking stall angle point, the embodiment of the present invention are not construed as limiting.

In the embodiment of the present invention, screening unit 403 can use traditional image processing method from the vehicle in aerial view A target parking stall frame is screened in the frame of position, and rejects other parking stall angles other than the target carriage parallactic angle point on the target parking stall frame Point.

In the embodiment of the present invention, the position that the first determination unit 405 first calculates the target carriage parallactic angle point in aerial view is sat Mark, then according to the mapping relations between the coordinate system of aerial view and world coordinate system, by the target parking stall in aerial view The position coordinates of angle point are converted into the position coordinates of target carriage parallactic angle point in world coordinate system, and by target carriage in world coordinate system The position coordinates of parallactic angle point are as the corresponding true location coordinate of target carriage parallactic angle point.

In the embodiment of the present invention, the second determination unit 406 is determined according to the corresponding true location coordinate of target carriage parallactic angle point The center in the region that target carriage parallactic angle point is constituted, the center in the region that target carriage parallactic angle point is constituted is as mesh The center of parking stall is marked, and target carriage parallactic angle point is placed in the corner of target parking stall.

As an alternative embodiment, first detection unit 402 can acquire a large amount of parking stall figure sample, according to Parking stall figure sample training obtains parking stall model, and above-mentioned aerial view is then imported the parking stall model and carries out convolution algorithm with life At various features, is detected from various features according to default feature and obtain parking stall frame and parking stall angle point.Implement above embodiment, The parking stall frame and parking stall angle point in aerial view are detected using deep learning algorithm, can he improve the precision of detection.

Further, as an alternative embodiment, first detection unit 402 can classify to feature, than Parking stall angle point is such as divided into 0 class, parking stall frame is divided into 1 class, above-mentioned aerial view is then imported into the parking stall model and carries out convolution 0 class therein is extracted as parking stall angle point to generate various features and 1 class therein is extracted conduct by operation Parking stall frame.Implement above embodiment, parking stall angle point is divided into 0 class and parking stall frame is divided into 1 class, can be improved computer Recognition speed, to improve the detection efficiency of parking stall.

As an alternative embodiment, screening unit 403 screens a target from the parking stall frame in aerial view Parking stall frame, comprising:

The width of each parking stall frame in aerial view is ranked up and counts the width of each parking stall frame by screening unit 403 The ranking of degree；

Screening unit 403 will rank the first corresponding parking stall frame as target parking stall frame.

As it can be seen that implementing stall testing apparatus described in Fig. 4, the ambient image of vehicle two sides is obtained by shooting, then The ambient image is converted into aerial view, a target parking stall frame and the target are screened from the parking stall frame in aerial view Target carriage parallactic angle point on the frame of parking stall obtains the corresponding true location coordinate of target carriage parallactic angle point and determining target carriage by calculating The corresponding target parking stall of position frame, can be improved the accuracy of parking stall measure, to improve the discrimination of parking stall.

Embodiment five

Referring to Fig. 5, Fig. 5 is the structural schematic diagram of another stall testing apparatus disclosed by the embodiments of the present invention.Wherein, Stall testing apparatus shown in fig. 5 is that stall testing apparatus as shown in Figure 4 advanced optimizes.With vehicle shown in Fig. 4 Level detecting apparatus compares, and stall testing apparatus shown in fig. 5 can also include:

Pretreatment unit 401 includes:

Distortion correction subelement 4011, for the inner parameter and external parameter according to the photographic device of vehicle, to shooting The ambient image for the vehicle two sides arrived carries out distortion correction, to obtain corresponding correcting image；

Perspective transform subelement 4012 carries out perspective transform to correcting image for the pixel according to correcting image, with Obtain corresponding aerial view.

First parameter acquiring unit 407, for the inside in distortion correction subelement 4011 according to the photographic device of vehicle Parameter and external parameter carry out distortion correction to the ambient image of the vehicle two sides taken, to obtain corresponding correcting image Before, the inner parameter and external parameter of the photographic device of vehicle are determined；

As an alternative embodiment, inner parameter of the distortion correction subelement 4011 according to the photographic device of vehicle And external parameter, distortion correction is carried out to the ambient image of the vehicle two sides taken, to obtain corresponding correcting image, packet It includes:

First determining module 40111, inner parameter and external ginseng for being determined according to the first parameter acquiring unit 407 Number, determines the first corresponding relationship between the ambient image of vehicle two sides and corresponding correcting image；

First preserving module 40112, for saving the first corresponding relationship to the first look-up table；

Distortion correction module 40113, for being carried out to the ambient image of the vehicle two sides taken according to the first look-up table Distortion correction, to obtain corresponding correcting image.

As an alternative embodiment, pixel of the perspective transform subelement 4012 according to correcting image, to correction Image carries out perspective transform, to obtain corresponding aerial view, comprising:

Second determining module 40121 determines correcting image and corresponding gets a bird's eye view view for the pixel according to correcting image The second corresponding relationship between figure；

Second preserving module 40122, for saving the second corresponding relationship to second look-up table；

Perspective transform module 40123, for perspective transform being carried out to correcting image, with acquisition pair according to second look-up table The aerial view answered.

Second detection unit 408, for, according to target parking stall frame, the mesh being obtained from the angle point of parking stall in acquiring unit 404 After target carriage parallactic angle point on mark parking stall frame and the first determination unit 405 is according to the target carriage parallactic angle point in aerial view Position coordinates, before determining the corresponding true location coordinate of target carriage parallactic angle point, detect target parking stall frame institute frame in area Whether there are obstacles in domain；

Marking unit 409, for detecting that there are obstacles for the region in target parking stall frame institute frame in second detection unit 408 When object, marking the region in target parking stall frame institute frame to be can not parking stall；

First determination unit 405, specifically for detecting the region in target parking stall frame institute frame in second detection unit 408 There is no when barrier, according to the position coordinates of the target carriage parallactic angle point in aerial view, determine that the target carriage parallactic angle point is corresponding True location coordinate.

As an alternative embodiment, second detection unit 408 can be multiple in the screening of the edge of target parking stall frame Region and the marginal density for detecting this multiple regions, using the marginal density in the maximum corresponding region of marginal density as object edge Density, then judges whether object edge density is greater than specified threshold, if so, showing the region there are barrier, label is single Member 409 marks the region in target parking stall frame institute frame to be can not parking stall；If not, the first determination unit 405 according to get a bird's eye view view The position coordinates of target carriage parallactic angle point in figure determine the corresponding true location coordinate of target carriage parallactic angle point.

Implement above embodiment, it can be determined that be in the region in the target parking stall frame institute frame of stall testing apparatus screening No there are barrier, avoiding parking collides barrier in the process, can be improved safety of parking.

As it can be seen that implementing stall testing apparatus described in Fig. 5, in determination compared with the stall testing apparatus described in Fig. 4 After target carriage parallactic angle point, whether there are obstacles for the region in detection target parking stall frame institute frame, if so, showing that the parking stall can not Stop, mark the target parking stall frame enclose in region be can not parking stall, can be avoided and bump against barrier during vehicle parking, Guarantee safety of parking.In addition, implementing stall testing apparatus described in Fig. 5, by the first look-up table pair for carrying distortion parameter The ambient images of the vehicle two sides taken carries out distortion correction and obtains correcting image, then the by carrying perspective transform coefficient Two look-up tables carry out perspective transform to correcting image and obtain aerial view, can be improved the formation speed of aerial view, improve vehicle Position detection efficiency.

Embodiment six

Referring to Fig. 6, Fig. 6 is the structural schematic diagram of another stall testing apparatus disclosed by the embodiments of the present invention.Wherein, Stall testing apparatus shown in fig. 6 is that stall testing apparatus as shown in Figure 4 advanced optimizes.With vehicle shown in Fig. 4 Level detecting apparatus compares, and stall testing apparatus shown in fig. 6 can also include:

Second parameter acquiring unit 410, for pretreatment unit 401 to the ambient images of the vehicle two sides taken into Row pretreatment, before obtaining corresponding aerial view, obtains the inner parameter and external parameter of the photographic device of vehicle；

Pretreatment unit 401 includes:

Determine subelement 4013, the inner parameter and external parameter for being determined according to the second parameter acquiring unit 410, really Determine the third corresponding relationship between the ambient image of vehicle two sides and corresponding aerial view；

Saving subunit 4014, for saving third corresponding relationship to third look-up table；

Get a bird's eye view transformation subelement 4015, for according to third look-up table, to the ambient images of the vehicle two sides taken into Row gets a bird's eye view transformation, to obtain corresponding aerial view.

As it can be seen that implementing stall testing apparatus described in Fig. 6, in determination compared with the stall testing apparatus described in Fig. 4 After target carriage parallactic angle point, whether there are obstacles for the region in detection target parking stall frame institute frame, if so, showing that the parking stall can not Stop, mark the target parking stall frame enclose in region be can not parking stall, can be avoided and bump against barrier during vehicle parking, Guarantee safety of parking.In addition, implementing stall testing apparatus described in Fig. 6, the third lookup of transformation coefficient is got a bird's eye view by carrying Table carries out the ambient image of the vehicle two sides taken to get a bird's eye view transformation acquisition aerial view, can be improved the generation of aerial view Speed improves parking stall measure efficiency.

Embodiment seven

Referring to Fig. 7, Fig. 7 is the structural schematic diagram of another stall testing apparatus disclosed by the embodiments of the present invention.Such as Fig. 7 Shown, which may include:

It is stored with the memory 701 of executable program code；

The processor 702 coupled with memory 701；

Wherein, processor 702 calls the executable program code stored in memory 701, and it is any one to execute FIG. 1 to FIG. 3 The method for detecting parking stalls of kind view-based access control model.

The embodiment of the present invention discloses a kind of computer readable storage medium, stores computer program, wherein the computer Program makes the method for detecting parking stalls of computer execution any one view-based access control model of FIG. 1 to FIG. 3.

The embodiment of the present invention is also disclosed a kind of using distribution platform, wherein using distribution platform for issuing computer journey Sequence product, wherein when computer program product is run on computers, so that computer executes such as the above each method embodiment In some or all of method step.

In various embodiments of the present invention, it should be appreciated that magnitude of the sequence numbers of the above procedures are not meant to execute suitable Successively, the execution sequence of each process should be determined by its function and internal logic the certainty of sequence, without coping with the embodiment of the present invention Implementation process constitutes any restriction.

Above-mentioned unit as illustrated by the separation member may or may not be physically separated, aobvious as unit The component shown may or may not be object unit, can be in one place, or may be distributed over multiple networks On unit.Some or all of units can be selected to achieve the purpose of the solution of this embodiment according to the actual needs.In addition, Each functional unit in various embodiments of the present invention can integrate in one processing unit, be also possible to the independent object of each unit Reason exists, and can also be integrated in one unit with two or more units.Above-mentioned integrated unit can both use hardware Form realize, can also realize in the form of software functional units.

In embodiment provided by the present invention, it should be appreciated that " B corresponding with A " indicates that B is associated with A, can be with according to A Determine B.It is also to be understood that determine that B is not meant to determine B only according to A according to A, it can also be according to A and/or other information Determine B.

Those of ordinary skill in the art will appreciate that all or part of the steps in the various methods of above-described embodiment is can It is completed with instructing relevant hardware by program, which can be stored in a computer readable storage medium, storage Medium include read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), programmable read only memory (Programmable Read-only Memory, PROM), erasable programmable is read-only deposits Reservoir (Erasable Programmable Read Only Memory, EPROM), disposable programmable read-only memory (One- Time Programmable Read-Only Memory, OTPROM), the electronics formula of erasing can make carbon copies read-only memory (Electrically-Erasable Programmable Read-Only Memory, EEPROM), CD-ROM (Compact Disc Read-Only Memory, CD-ROM) or other disc memories, magnetic disk storage, magnetic tape storage or can For carrying or any other computer-readable medium of storing data.

Detailed Jie has been carried out to the method for detecting parking stalls and device of a kind of view-based access control model disclosed by the embodiments of the present invention above It continues, used herein a specific example illustrates the principle and implementation of the invention, and the explanation of above embodiments is only It is to be used to help understand method and its core concept of the invention；At the same time, for those skilled in the art, according to this hair Bright thought, there will be changes in the specific implementation manner and application range, in conclusion the content of the present specification should not manage Solution is limitation of the present invention.

Claims

1. a kind of method for detecting parking stalls of view-based access control model characterized by comprising

According to target parking stall frame, from the target carriage parallactic angle point obtained in the parking stall angle point on the frame of the target parking stall；

According to the position coordinates of the target carriage parallactic angle point in the aerial view, determine that the target carriage parallactic angle point is corresponding True location coordinate；

According to the corresponding true location coordinate of the target carriage parallactic angle point, the corresponding target parking stall of target parking stall frame is determined.

2. the method according to claim 1, wherein the ambient image of the described pair of vehicle two sides taken carries out Pretreatment, to obtain corresponding aerial view, comprising:

According to the inner parameter and external parameter of the photographic device of vehicle, to the ambient images of the vehicle two sides taken into Line distortion correction, to obtain corresponding correcting image；

According to the pixel of the correcting image, perspective transform is carried out to the correcting image, to obtain corresponding aerial view.

3. according to the method described in claim 2, it is characterized in that, the photographic device according to vehicle inner parameter and External parameter carries out distortion correction to the ambient images of the vehicle two sides taken, with obtain corresponding correcting image it Before, the method also includes:

Obtain the inner parameter and external parameter of the photographic device of vehicle；The inner parameter includes at least the photographic device Focal length, Pixel Dimensions and distortion parameter；The external parameter includes at least height and institute of the photographic device relative to ground State the spin matrix of photographic device；

The inner parameter and external parameter of the photographic device according to vehicle, to the environment map of the vehicle two sides taken As carrying out distortion correction, to obtain corresponding correcting image, comprising:

According to the inner parameter and the external parameter, determine the vehicle two sides ambient image and corresponding correcting image Between the first corresponding relationship；First corresponding relationship includes the pixel of the ambient image of the vehicle two sides taken Corresponding relationship between the pixel of the correcting image；

First corresponding relationship is saved to the first look-up table；

According to first look-up table, distortion correction is carried out to the ambient image of the vehicle two sides taken, to obtain State correcting image；

The pixel according to the correcting image carries out perspective transform to the correcting image, to obtain corresponding get a bird's eye view View, comprising:

According to the pixel of the correcting image, the second corresponding pass between the correcting image and corresponding aerial view is determined System；Second corresponding relationship includes the corresponding pass between the pixel of the correcting image and the pixel of the aerial view System；

Second corresponding relationship is saved to second look-up table；

4. the method according to claim 1, wherein the ambient image of the vehicle two sides taken at described Dui into Row pretreatment, before obtaining corresponding aerial view, the method also includes:

And the ambient image of the described pair of vehicle two sides taken pre-processes, and to obtain corresponding aerial view, packet It includes:

According to the inner parameter and the external parameter, determine the vehicle two sides ambient image and corresponding aerial view Between third corresponding relationship；The third corresponding relationship includes the pixel of the ambient image of the vehicle two sides taken Corresponding relationship between the pixel of the aerial view；

The third corresponding relationship is saved to third look-up table；

According to the third look-up table, the ambient image of the vehicle two sides taken is carried out getting a bird's eye view transformation, to obtain State aerial view.

5. method according to any one of claims 1 to 4, which is characterized in that described according to target parking stall frame, from It is obtained in the parking stall angle point after the target carriage parallactic angle point on the frame of the target parking stall and described according to the aerial view In the target carriage parallactic angle point position coordinates, before determining the corresponding true location coordinate of the target carriage parallactic angle point, institute State method further include:

If region in target parking stall frame institute frame there are barrier, marks the region in target parking stall frame institute frame to be It can not parking stall；

If barrier is not present in the region in target parking stall frame institute frame, the institute according in the aerial view is executed The position coordinates for stating target carriage parallactic angle point, the step of determining the target carriage parallactic angle point corresponding true location coordinate.

6. a kind of stall testing apparatus characterized by comprising

Pretreatment unit is pre-processed for the ambient image to the vehicle two sides taken, corresponding gets a bird's eye view view to obtain Figure；

Acquiring unit is used for according to target parking stall frame, from the mesh obtained on the frame of the target parking stall in the parking stall angle point Mark parking stall angle point；

First determination unit, for the position coordinates according to the target carriage parallactic angle point in the aerial view, determine described in The corresponding true location coordinate of target carriage parallactic angle point；

Second determination unit, for determining the target parking stall according to the corresponding true location coordinate of the target carriage parallactic angle point The corresponding target parking stall of frame.

7. stall testing apparatus according to claim 6, which is characterized in that the pretreatment unit includes:

Distortion correction subelement, for the inner parameter and external parameter according to the photographic device of vehicle, described in taking The ambient image of vehicle two sides carries out distortion correction, to obtain corresponding correcting image；

Perspective transform subelement carries out perspective transform to the correcting image for the pixel according to the correcting image, with Obtain corresponding aerial view.

8. stall testing apparatus according to claim 7, which is characterized in that the stall testing apparatus further include:

First parameter acquiring unit, in the distortion correction subelement according to the inner parameter of the photographic device of vehicle and outer Portion's parameter carries out distortion correction to the ambient image of the vehicle two sides taken, before obtaining corresponding correcting image, Obtain the inner parameter and external parameter of the photographic device of vehicle；The inner parameter includes at least the coke of the photographic device Away from, Pixel Dimensions and distortion parameter；The external parameter includes at least the photographic device relative to the height on ground and described The spin matrix of photographic device；

The distortion correction subelement includes:

First determining module, for determining the environment map of the vehicle two sides according to the inner parameter and the external parameter The first corresponding relationship between picture and corresponding correcting image；First corresponding relationship includes the vehicle two sides taken The pixel of ambient image and the pixel of the correcting image between corresponding relationship；

Distortion correction module, for being carried out to the ambient image of the vehicle two sides taken according to first look-up table Distortion correction, to obtain the correcting image；

The perspective transform subelement includes:

Second determining module determines the correcting image and corresponding gets a bird's eye view view for the pixel according to the correcting image The second corresponding relationship between figure；Second corresponding relationship includes the pixel and the aerial view of the correcting image Corresponding relationship between pixel；

Second preserving module, for saving second corresponding relationship to second look-up table；

Perspective transform module, for perspective transform being carried out to the correcting image, described in obtaining according to the second look-up table Aerial view.

9. stall testing apparatus according to claim 6, which is characterized in that the stall testing apparatus further include:

Second parameter acquiring unit, for being located in advance in ambient image of the pretreatment unit to the vehicle two sides taken Reason, before obtaining corresponding aerial view, obtains the inner parameter and external parameter of the photographic device of vehicle；The internal ginseng Number includes at least focal length, Pixel Dimensions and the distortion parameter of the photographic device；The external parameter includes at least the camera shooting Device is relative to the height on ground and the spin matrix of the photographic device；

The pretreatment unit includes:

Subelement is determined, for determining the ambient image of the vehicle two sides according to the inner parameter and the external parameter Third corresponding relationship between corresponding aerial view；The third corresponding relationship includes the vehicle two sides taken Corresponding relationship between the pixel of ambient image and the pixel of the aerial view；

Get a bird's eye view transformation subelement, for according to the third look-up table, to the ambient images of the vehicle two sides taken into Row gets a bird's eye view transformation, to obtain the aerial view.

10. according to the described in any item stall testing apparatus of claim 6 to 9, which is characterized in that the stall testing apparatus is also Include:

Second detection unit, for, according to target parking stall frame, obtaining institute from the parking stall angle point in the acquiring unit It states after the target carriage parallactic angle point on the frame of target parking stall and first determination unit is according in the aerial view The position coordinates of target carriage parallactic angle point detect the mesh before determining the corresponding true location coordinate of the target carriage parallactic angle point Whether there are obstacles in region in mark parking stall frame institute frame；

Marking unit, for detecting that there are barriers for the region in target parking stall frame institute frame in the second detection unit When, marking the region in target parking stall frame institute frame to be can not parking stall；

First determination unit, specifically for detecting the area in target parking stall frame institute frame in the second detection unit When barrier is not present in domain, according to the position coordinates of the target carriage parallactic angle point in the aerial view, the target is determined The corresponding true location coordinate of parking stall angle point.