CN113053164A - Parking space identification method using look-around image - Google Patents
Parking space identification method using look-around image Download PDFInfo
- Publication number
- CN113053164A CN113053164A CN202110514586.3A CN202110514586A CN113053164A CN 113053164 A CN113053164 A CN 113053164A CN 202110514586 A CN202110514586 A CN 202110514586A CN 113053164 A CN113053164 A CN 113053164A
- Authority
- CN
- China
- Prior art keywords
- parking space
- image
- straight line
- template
- point
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/14—Traffic control systems for road vehicles indicating individual free spaces in parking areas
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/16—Anti-collision systems
- G08G1/168—Driving aids for parking, e.g. acoustic or visual feedback on parking space
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a parking space identification method by using a look-around image, which comprises the steps of preprocessing the look-around image, determining the position of a parking space angular point by adopting a double constraint condition of combining linear detection and angular point detection, and identifying the type and the position of a parking space by using a template matching method. The invention reduces the error recognition rate of parking space recognition by means of double constraint conditions, and can be applied to advanced driving auxiliary systems such as automatic parking and the like.
Description
Technical Field
The invention relates to the technical field of parking space identification, in particular to a parking space identification method by using a look-around image.
Background
Under the condition of the existing all-round view image, the parking space is identified by only using one constraint condition, so that the error identification and the non-identification can be caused. There is a need for a parking space recognition method capable of improving recognition accuracy to solve the above problems.
Disclosure of Invention
Aiming at the defects of the prior art, the invention aims to provide a parking space identification method by using a look-around image.
In order to achieve the purpose, the invention adopts the following technical scheme:
a parking space identification method using a look-around image comprises the following specific processes:
s1, preprocessing the image;
s2, determining the inflection point of the parking space:
s2.1, firstly, detecting a straight line of a parking space by using a Hough line detection method, acquiring the slope and intercept of the straight line, and packaging the detected straight line and the slope and intercept data thereof into a data structure of index + data;
s2.2, classifying the straight lines according to the slope and the intercept, and combining the straight line segments with the slopes and the intercepts within a set error range;
because the parking space straight line has width, only the parallel straight line and the straight line with intercept within the set range can be regarded as the straight line of the parking space, and the straight line which does not accord with the width characteristic is removed according to the limiting condition;
because the slope of the straight line of the parking space only meets three possibilities of 60 degrees, 90 degrees and 120 degrees, the slope which does not meet the conditions and the corresponding intercept are removed according to the limiting conditions;
after the slope and intercept of the straight line of the parking space are combined and screened, primarily establishing the inflection point of the parking space, wherein the initially established inflection point of the parking space is the intersection point of the median lines of the straight line of the parking space;
s2.3, screening out the parking space lines which do not accord with the limitation condition according to the parking space width and the limitation condition that one parking space needs two parallel straight lines;
s2.4, eliminating false detection by means of geometric constraint of the parking space, wherein the geometric constraint of the parking space is the length-width ratio of the parking space set in the industry standard, reserving the parking space which meets the constraint condition, and eliminating the parking space which does not meet the constraint condition;
s2.5, after primary screening of the parking space inflection point is completed, improving the confidence coefficient of identification by combining a FAST angular point detection algorithm, wherein the basis of angular point detection of the algorithm is to judge the relation between a point to be detected and surrounding points, but the inflection point detected by the algorithm comprises a true point and an interference point, so that the position relation between the inflection point and a straight line is utilized, the Euclidean distance is calculated, and the parking space inflection point is finally determined;
s3, recognizing the parking space by using template matching:
firstly, extracting an interested area image according to the position of the parking space inflection point determined in the step S2; then, thinning the image of the region of interest;
then, a template element set which has the same size as the image of the region of interest and is used for template matching is made; and then matching each element in the template element set with the refined region-of-interest image, matching the element in the template element set with the refined image again by rotating the element by a fixed angle until the best matching degree is output, and finally identifying the type and the position of the parking space.
Further, the specific process of step S1 is:
s1.1, taking a central area of a panoramic image as an automobile image, establishing an automobile mask image, and carrying out bitwise AND operation on the automobile mask image and an original image to finally obtain a color image with effective information;
s1.2, graying the color image obtained in the step S1.1 to obtain a gray image;
and S1.3, denoising the gray level image obtained in the step S1.2 to finally obtain an image subjected to denoising treatment.
Furthermore, in step S1.2, graying is performed by a weighted average method; the formula for the weighted average is:
Grayi,j＝γ*Redi,j+Ψ*Greeni,j+Ω*Bluei,jwherein γ + Ψ + Ω ═ 1;
Grayi,jthe pixel value of a pixel point with an abscissa of i and an ordinate of j in the processed gray level image is Redi,jThe Green is the pixel value of a pixel point with the abscissa of i and the ordinate of j in a red channel of the color imagei,jIs the pixel value of a pixel point with the abscissa of i and the ordinate of j in a green channel of a color image, Bluei,jThe pixel values of pixel points with an abscissa of i and an ordinate of j in a blue channel of the color image are obtained, gamma, psi and omega are respectively weight coefficients of red, green and blue channels, and the sum of the weight coefficients is 1.
Further, in step S1.3, noise reduction is achieved using the following noise reduction formula:
Dm,n＝Sm-1,n-1*h0,0+Sm-1,n*h0,1+…+Sm+1,n+1*h2,2；
h0,0,h0,1,……h2,2respective parameters of the filter kernel function, S, of size 3 x 3, respectivelym-1,n-1，Sm-1,n，……,Sm+1,n+1The pixel values D of pixel points in eight directions with the abscissa m and the ordinate n of the original image as the centerm,nIs the processed pixel value.
Furthermore, in step S2.3, the influence of the illumination on the linear width is taken into consideration, the illumination condition at this time is determined through the calculation of the histogram, adjustment is performed according to the illumination condition, and when the illumination is weak, a constraint condition is added to screen out the vehicle location line which does not meet the constraint condition, so as to improve the identification capability of the parking space.
Further, in step S3, the matching degree Sim between the template image and the thinned imageTRUsing a formula
To calculate; simTRExpressed as the similarity of the template image and the refined image; phi is aTRepresents the average value of the pixel values of the template image, phiRRepresenting the mean value, ζ, of the pixel values of the refined imageTRRepresenting the covariance, ζ, of the template and refined imagesTRepresents the mean square error, ζ, of the template image pixelsRRepresenting the mean square error of the pixels of the refined image, wherein m and n are two constants; if the matching degree function value SimTRIf the value is larger than the threshold xi, the matching is successful; if the matching degree function value SimTRLess than the threshold ξ, the match fails.
The invention has the beneficial effects that: according to the invention, firstly, the panoramic image is preprocessed, secondly, the position of the parking space angular point is determined by adopting a double constraint condition of combining linear detection and angular point detection, and finally, the type and the position of the parking space are identified by adopting a template matching method. The invention reduces the error recognition rate of parking space recognition by means of double constraint conditions, and can be applied to advanced driving auxiliary systems such as automatic parking and the like.
Drawings
Fig. 1 is a schematic diagram of a specific process of determining a parking space inflection point in the method according to the embodiment of the present invention;
fig. 2 is a schematic diagram of a specific process of template matching in the method according to the embodiment of the present invention.
Detailed Description
The present invention will be further described with reference to the accompanying drawings, and it should be noted that the present embodiment is based on the technical solution, and the detailed implementation and the specific operation process are provided, but the protection scope of the present invention is not limited to the present embodiment.
The embodiment provides a parking space identification method using a look-around image, which specifically comprises the following processes:
and S1, preprocessing the image.
S1.1, in the conventional all-around image, not all regions need to be processed. The central area of the all-round view image is an automobile image, the part is meaningless in subsequent processing and is only for beauty, so that the information of the area is removed, the calculation amount of subsequent operation can be effectively reduced, and the subsequent parking space recognition rate can be effectively improved. Therefore, in this embodiment, a car mask image is created, and the car mask image and the original image are bitwise and operated, so that a color image with effective information can be obtained finally.
S1.2, in order to reduce the processing complexity, graying the color image obtained in the step S1.1 to obtain a grayscale image. This may be done using a weighted average method. The formula for the weighted average is:
Grayi,j＝γ*Redi,j+Ψ*Greeni,j+Ω*Bluei,jwherein γ + Ψ + Ω ═ 1.
Grayi,jThe pixel value of a pixel point with an abscissa of i and an ordinate of j in the processed gray level image is Redi,jThe Green is the pixel value of a pixel point with the abscissa of i and the ordinate of j in a red channel of the color imagei,jIs the pixel value of a pixel point with the abscissa of i and the ordinate of j in a green channel of a color image, Bluei,jThe abscissa of the blue channel of the color image is i,The pixel value of the pixel point with the ordinate j, γ, Ψ, Ω are the weighting coefficients of the red, green, and blue channels, respectively, and the total sum is 1.
And S1.3, the environment near the parking space is complex and inevitably generates noise, so that the identification of the subsequent parking space is influenced, and therefore, the embodiment performs noise reduction on the gray-scale image obtained in the step S1.2, and finally obtains the image subjected to noise reduction processing.
The formula of noise reduction is:
Dm,n＝Sm-1,n-1*h0,0+Sm-1,n*h0,1+…+Sm+1,n+1*h2,2；
h0,0,h0,1,……h2,2respective parameters of the filter kernel function, S, of size 3 x 3, respectivelym-1,n-1，Sm-1,n，……,Sm+1,n+1The pixel values D of pixel points in eight directions with the abscissa m and the ordinate n of the original image as the centerm,nIs the processed pixel value.
And S2, determining the inflection point of the parking space.
The method for accurately positioning the parking space needs to acquire the position of a parking space straight line and the position of an inflection point, and the inflection point of the parking space is determined by adopting a method of double constraints of straight line detection and angular point detection. A normal parking space is defined by a straight line and an inflection point, so that the position of the straight line and the position of the inflection point need to be clearly known when a parking space is located. The present embodiment uses a data structure consisting of two parts, data and index. Through the data structure, double-sequencing operation can be carried out on the slope and the intercept of the straight line, and initial screening of the slope is completed. As shown in fig. 1, the specific process is as follows:
s2.1, firstly, detecting a straight line of a parking space by using a Hough line detection method, acquiring the slope and intercept of the straight line, and packaging the detected straight line and the slope and intercept data thereof into a data structure of index + data.
S2.2, in the detected straight line result, an original complete straight line may be divided into several line segments due to some factors, so the present embodiment classifies the straight lines according to the slope and the intercept. The slope and intercept of the straight line are unique, so that the slope of the line segments in the line segment set belonging to the same straight line is approximate, and on the basis, the intercept is combined to judge whether the line segments belong to the same straight line or not so as to complete extension and combination of the straight line segments, and particularly, the straight line segments with the slopes and the intercepts within a set error range are combined together.
Because the parking space straight line has width, the removal of useless straight lines is facilitated, and only straight lines which are parallel and have intercept within a set range can be regarded as straight lines of the parking space. And removing the straight lines which do not accord with the width characteristic according to the limiting condition.
Because the slope of the straight line of the parking space only meets three possibilities of 60 degrees, 90 degrees and 120 degrees, the slope and the corresponding intercept which do not meet the conditions are removed according to the limiting conditions.
The slope and intercept of the straight line of the parking space can be merged and screened through the process, so that the primary establishment of the inflection point of the parking space is realized, and the initially established inflection point of the parking space is the intersection point of the median line of the straight line of the parking space.
And S2.3, according to the industry standard of the parking space on the road in China, the parking space has three types of parallel parking space, inclined parking space and vertical parking space. The three types of parking spaces have the characteristics of two parallel straight lines, the size of each parking space is 6m × 2.5m, and the width of each parking space straight line is 10cm, so that a parking space line which does not accord with the restriction condition is screened according to the width of each parking space and the restriction condition that one parking space needs two parallel straight lines. When the screening is carried out, the influence of the illumination on the linear width is considered, the illumination condition at the moment is judged through the calculation of the histogram, the adjustment is carried out according to the illumination condition, and when the illumination is weak, the constraint condition is added so as to improve the identification capability of the parking space.
S2.4, finally, eliminating false detection by means of geometric constraint of the parking space, wherein the geometric constraint of the parking space is the length-width ratio of the parking space set in the industry standard, in the detection process, the same-shaped patterns may appear, but the corresponding ratios are different, the patterns meeting the constraint condition need to be reserved, and the patterns not meeting the constraint condition are eliminated.
S2.5, after the initial screening of the parking space inflection point is completed, the confidence coefficient of identification is improved by combining a FAST angular point detection algorithm, the basis of angular point detection of the algorithm is to judge the relation between a point to be detected and surrounding points, but the inflection point detected by the algorithm comprises a real point and an interference point, so that the position relation between the inflection point and a straight line is utilized, the Euclidean distance is calculated, and the parking space inflection point is finally determined.
And S3, identifying the parking space by utilizing template matching.
In this embodiment, a template matching method is used to accurately identify the type and position of the inflection point of the parking space, and further remove the misidentification caused by the previous stage, where the specific process is shown in fig. 2:
s3.1, firstly, extracting an interested area image according to the position of the inflection point of the parking space determined in the step S2. And then, carrying out thinning operation on the image of the region of interest, wherein the thinning operation is to improve the efficiency and the precision of template matching and accurately obtain structural information near an inflection point. Then, a template element set for template matching is made to have the same size as the region-of-interest image. Each element in the template element set is then matched to the refined region of interest image. And matching the elements in the template element set with the refined image again by rotating the elements by a fixed angle until the best matching degree is output, and finally identifying the type and the position of the parking space.
Matching degree Sim between template image and refined imageTRCan adopt a formula
To calculate. SimTRExpressed as the similarity of the template image and the refined image. Phi is aTRepresents the average value of the pixel values of the template image, phiRRepresenting the mean value, ζ, of the pixel values of the refined imageTRRepresenting the covariance, ζ, of the template and refined imagesTRepresents the mean square error, ζ, of the template image pixelsRRepresentation refinementThe mean square error of an image pixel, m, n, is two constants. If the matching degree function value SimTRAnd if the value is larger than the threshold xi, the matching is successful. If the matching degree function value SimTRLess than the threshold ξ, the match fails.
Various corresponding changes and modifications can be made by those skilled in the art based on the above technical solutions and concepts, and all such changes and modifications should be included in the protection scope of the present invention.
Claims (6)
1. The utility model provides a parking stall identification method of utilizing look around image which characterized in that, the concrete process is:
s1, preprocessing the image;
s2, determining the inflection point of the parking space:
s2.1, firstly, detecting a straight line of a parking space by using a Hough line detection method, acquiring the slope and intercept of the straight line, and packaging the detected straight line and the slope and intercept data thereof into a data structure of index + data;
s2.2, classifying the straight lines according to the slope and the intercept, and combining the straight line segments with the slopes and the intercepts within a set error range;
because the parking space straight line has width, only the parallel straight line and the straight line with intercept within the set range can be regarded as the straight line of the parking space, and the straight line which does not accord with the width characteristic is removed according to the limiting condition;
because the slope of the straight line of the parking space only meets three possibilities of 60 degrees, 90 degrees and 120 degrees, the slope which does not meet the conditions and the corresponding intercept are removed according to the limiting conditions;
after the slope and intercept of the straight line of the parking space are combined and screened, primarily establishing the inflection point of the parking space, wherein the initially established inflection point of the parking space is the intersection point of the median lines of the straight line of the parking space;
s2.3, screening out the parking space lines which do not accord with the limitation condition according to the parking space width and the limitation condition that one parking space needs two parallel straight lines;
s2.4, eliminating false detection by means of geometric constraint of the parking space, wherein the geometric constraint of the parking space is the length-width ratio of the parking space set in the industry standard, reserving the parking space which meets the constraint condition, and eliminating the parking space which does not meet the constraint condition;
s2.5, after primary screening of the parking space inflection point is completed, improving the confidence coefficient of identification by combining a FAST angular point detection algorithm, wherein the basis of angular point detection of the algorithm is to judge the relation between a point to be detected and surrounding points, but the inflection point detected by the algorithm comprises a true point and an interference point, so that the position relation between the inflection point and a straight line is utilized, the Euclidean distance is calculated, and the parking space inflection point is finally determined;
s3, recognizing the parking space by using template matching:
firstly, extracting an interested area image according to the position of the parking space inflection point determined in the step S2; then, thinning the image of the region of interest;
then, a template element set which has the same size as the image of the region of interest and is used for template matching is made; and then matching each element in the template element set with the refined region-of-interest image, matching the element in the template element set with the refined image again by rotating the element by a fixed angle until the best matching degree is output, and finally identifying the type and the position of the parking space.
2. The method according to claim 1, wherein the specific process of step S1 is as follows:
s1.1, taking a central area of a panoramic image as an automobile image, establishing an automobile mask image, and carrying out bitwise AND operation on the automobile mask image and an original image to finally obtain a color image with effective information;
s1.2, graying the color image obtained in the step S1.1 to obtain a gray image;
and S1.3, denoising the gray level image obtained in the step S1.2 to finally obtain an image subjected to denoising treatment.
3. The method according to claim 2, characterized in that in step S1.2, graying is performed by a weighted average method; the formula for the weighted average is:
Grayi,j＝γ*Redi,j+Ψ*Greeni,j+Ω*Bluei,jwherein γ + Ψ + Ω ═ 1;
Grayi,jthe pixel value of a pixel point with an abscissa of i and an ordinate of j in the processed gray level image is Redi,jThe Green is the pixel value of a pixel point with the abscissa of i and the ordinate of j in a red channel of the color imagei,jIs the pixel value of a pixel point with the abscissa of i and the ordinate of j in a green channel of a color image, Bluei,jThe pixel values of pixel points with an abscissa of i and an ordinate of j in a blue channel of the color image are obtained, gamma, psi and omega are respectively weight coefficients of red, green and blue channels, and the sum of the weight coefficients is 1.
4. A method according to claim 2, characterized in that in step S1.3, the noise reduction is implemented using the following noise reduction formula:
Dm,n＝Sm-1,n-1*h0,0+Sm-1,n*h0,1+…+Sm+1,n+1*h2,2；
h0,0,h0,1,……h2,2respective parameters of the filter kernel function, S, of size 3 x 3, respectivelym-1,n-1，Sm-1,n，……,Sm+1,n+1The pixel values D of pixel points in eight directions with the abscissa m and the ordinate n of the original image as the centerm,nIs the processed pixel value.
5. The method according to claim 1, wherein in step S2.3, considering the influence of the illumination on the linear width, the illumination condition at that time is determined by the calculation of the histogram, and the adjustment is performed according to the illumination condition, and when the illumination is weak, the constraint condition is added to screen out the vehicle location lines that do not meet the constraint condition, so as to improve the recognition capability of the parking space.
6. The method according to claim 1, wherein in step S3, the matching degree Sim between the template image and the refined imageTRUsing a formula
To calculate; simTRExpressed as the similarity of the template image and the refined image; phi is aTRepresents the average value of the pixel values of the template image, phiRRepresenting the mean value, ζ, of the pixel values of the refined imageTRRepresenting the covariance, ζ, of the template and refined imagesTRepresents the mean square error, ζ, of the template image pixelsRRepresenting the mean square error of the pixels of the refined image, wherein m and n are two constants; if the matching degree function value SimTRIf the value is larger than the threshold xi, the matching is successful; if the matching degree function value SimTRLess than the threshold ξ, the match fails.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110514586.3A CN113053164A (en) | 2021-05-11 | 2021-05-11 | Parking space identification method using look-around image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110514586.3A CN113053164A (en) | 2021-05-11 | 2021-05-11 | Parking space identification method using look-around image |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113053164A true CN113053164A (en) | 2021-06-29 |
Family
ID=76518254
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110514586.3A Pending CN113053164A (en) | 2021-05-11 | 2021-05-11 | Parking space identification method using look-around image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113053164A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113593296A (en) * | 2021-07-26 | 2021-11-02 | 广州小鹏自动驾驶科技有限公司 | Parking space detection method and device, vehicle and medium |
| CN114821540A (en) * | 2022-05-27 | 2022-07-29 | 禾多科技（北京）有限公司 | Parking space detection method and device, electronic equipment and computer readable medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109299686A (en) * | 2018-09-14 | 2019-02-01 | 深圳市灵动飞扬科技有限公司 | A kind of parking stall recognition methods, device, equipment and medium |
| CN109435942A (en) * | 2018-10-31 | 2019-03-08 | 合肥工业大学 | A kind of parking stall line parking stall recognition methods and device based on information fusion |
| CN109649384A (en) * | 2019-02-15 | 2019-04-19 | 华域汽车系统股份有限公司 | A kind of parking assistance method |
| CN110472580A (en) * | 2019-08-16 | 2019-11-19 | 北京茵沃汽车科技有限公司 | The method, apparatus of detection parking stall based on full-view image, storage medium |
| KR20200068045A (en) * | 2018-11-26 | 2020-06-15 | (주)캠시스 | Around view monitoring system and method for detecting parking line with deep learning |
| CN112598922A (en) * | 2020-12-07 | 2021-04-02 | 安徽江淮汽车集团股份有限公司 | Parking space detection method, device, equipment and storage medium |
-
2021
- 2021-05-11 CN CN202110514586.3A patent/CN113053164A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109299686A (en) * | 2018-09-14 | 2019-02-01 | 深圳市灵动飞扬科技有限公司 | A kind of parking stall recognition methods, device, equipment and medium |
| CN109435942A (en) * | 2018-10-31 | 2019-03-08 | 合肥工业大学 | A kind of parking stall line parking stall recognition methods and device based on information fusion |
| KR20200068045A (en) * | 2018-11-26 | 2020-06-15 | (주)캠시스 | Around view monitoring system and method for detecting parking line with deep learning |
| CN109649384A (en) * | 2019-02-15 | 2019-04-19 | 华域汽车系统股份有限公司 | A kind of parking assistance method |
| CN110472580A (en) * | 2019-08-16 | 2019-11-19 | 北京茵沃汽车科技有限公司 | The method, apparatus of detection parking stall based on full-view image, storage medium |
| CN112598922A (en) * | 2020-12-07 | 2021-04-02 | 安徽江淮汽车集团股份有限公司 | Parking space detection method, device, equipment and storage medium |
Non-Patent Citations (4)
| Title |
|---|
| JINDONG ZHANG 等: "An improved parking space recognition algorithm based on panoramic vision", 《MULTIMEDIA TOOLS AND APPLICATIONS》 * |
| 房翔: "基于车载环视系统的车位识别算法研究", 《中国优秀博硕士学位论文全文数据库(硕士)工程科技Ⅱ辑》 * |
| 郭剑鹰 等: "基于多视觉传感器的泊车辅助系统设计", 《传动技术》 * |
| 韩雨松: "基于双目视觉的车库泊车位检测研究与实现", 《中国优秀硕士学位论文全文数据库(工程科技Ⅱ辑)》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113593296A (en) * | 2021-07-26 | 2021-11-02 | 广州小鹏自动驾驶科技有限公司 | Parking space detection method and device, vehicle and medium |
| CN114821540A (en) * | 2022-05-27 | 2022-07-29 | 禾多科技（北京）有限公司 | Parking space detection method and device, electronic equipment and computer readable medium |
| CN114821540B (en) * | 2022-05-27 | 2023-03-24 | 禾多科技（北京）有限公司 | Parking space detection method and device, electronic equipment and computer readable medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108564814B (en) | Image-based parking lot parking space detection method and device | |
| CN106778551B (en) | Method for identifying highway section and urban road lane line | |
| CN107492076B (en) | Method for suppressing vehicle shadow interference in expressway tunnel scene | |
| CN112819094A (en) | Target detection and identification method based on structural similarity measurement | |
| CN109460722B (en) | Intelligent license plate recognition method | |
| CN107895151A (en) | Method for detecting lane lines based on machine vision under a kind of high light conditions | |
| CN109886168B (en) | Ground traffic sign identification method based on hierarchy | |
| CN113053164A (en) | Parking space identification method using look-around image | |
| CN107563301A (en) | Red signal detection method based on image processing techniques | |
| Babbar et al. | A new approach for vehicle number plate detection | |
| CN112528868B (en) | Illegal line pressing judgment method based on improved Canny edge detection algorithm | |
| CN109858438B (en) | Lane line detection method based on model fitting | |
| CN114511770A (en) | Road sign plate identification method | |
| CN116152115B (en) | Garbage image denoising processing method based on computer vision | |
| CN117036341A (en) | Pavement crack detection method based on image processing | |
| CN113537211A (en) | Deep learning license plate frame positioning method based on asymmetric IOU | |
| CN111652033A (en) | Lane line detection method based on OpenCV | |
| CN109063669B (en) | Bridge area ship navigation situation analysis method and device based on image recognition | |
| CN111401364A (en) | License plate positioning algorithm based on combination of color features and template matching | |
| CN115018785A (en) | Hoisting steel wire rope tension detection method based on visual vibration frequency identification | |
| FAN et al. | Robust lane detection and tracking based on machine vision | |
| CN107977608B (en) | Method for extracting road area of highway video image | |
| CN108520252B (en) | Road sign identification method based on generalized Hough transform and wavelet transform | |
| CN110516666B (en) | License plate positioning method based on combination of MSER and ISODATA | |
| Chen et al. | A novel lane departure warning system for improving road safety |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210629 |