CN105809131B - A kind of method and system carrying out parking stall water detection based on image processing techniques - Google Patents
A kind of method and system carrying out parking stall water detection based on image processing techniques Download PDFInfo
- Publication number
- CN105809131B CN105809131B CN201610129872.7A CN201610129872A CN105809131B CN 105809131 B CN105809131 B CN 105809131B CN 201610129872 A CN201610129872 A CN 201610129872A CN 105809131 B CN105809131 B CN 105809131B
- Authority
- CN
- China
- Prior art keywords
- ponding
- image
- module
- complexity
- parking stall
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/56—Context or environment of the image exterior to a vehicle by using sensors mounted on the vehicle
- G06V20/58—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10016—Video; Image sequence
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a kind of method and system for carrying out parking stall water detection based on image processing techniques, this method includes outdoor scene ponding model training on daytime；The calibration of backsight fish eye lens, saves internal reference matrix and distortion factor matrix；The video flowing for obtaining the acquisition of backsight fish eye lens, makees distortion correction to image；Detection zone is set, color notation conversion space is made to image in region, analyzes pixel color characteristics and saturation degree, retains the pixel for meeting given threshold；Morphological scale-space and connected component analysis, the blob image block for meeting area threshold make profile analysis of complexity；Water detection result is obtained after comparison.System is made of ponding model training module, fish eye lens demarcating module, image distortion correction module, color characteristics analysis module, profile analysis of complexity module, feature extraction and model comparison module.The present invention can quickly and efficiently identify parking stall ponding, more be able to satisfy the intelligent demand of parking of user.
Description
Technical field
The present invention relates to field of automobile safety, specifically a kind of side that parking stall water detection is carried out based on image processing techniques
Method and system.
Background technique
Automotive engineering is maked rapid progress, and the concepts such as unmanned, car networking, intelligent automobile have become hot topic instantly,
The technologies such as big data related to this, artificial intelligence, image procossing increasingly become the hot and difficult issue of the area research.
Intelligent parking technology based on video image is a kind of to carry out feature to acquired video flowing using image algorithm
Analysis finally guides vehicle automatic stopping to enter the intellectual technology of position.The technology has important meaning to the research in unmanned field
Justice.These technologies the modes such as are mainly detected and are positioned by parking stall line and realize at present, and barrier correlation is believed in shorter mention parking stall
The analysis of breath is especially not directed to the detection and analysis of parking stall internal water accumulation situation.
Summary of the invention
The purpose of the present invention is to provide it is a kind of based on image processing techniques carry out parking stall water detection method and system,
It can quickly and efficiently identify parking stall ponding, more be able to satisfy the intelligent demand of parking of user.
To achieve the above object, the invention provides the following technical scheme:
A method of parking stall water detection is carried out based on image processing techniques, comprising the following steps:
1) outdoor scene ponding model training on daytime；
2) backsight fish eye lens is demarcated, and saves internal reference matrix and distortion factor matrix；
3) video flowing for obtaining the acquisition of backsight fish eye lens, makees distortion correction to image；
4) detection zone is set, color notation conversion space is made to image in region, analyzes pixel color characteristics and saturation degree,
Retain the pixel for meeting given threshold；
5) Morphological scale-space and connected component analysis, meet area threshold blob (binary large object, two
System blob) image block makees profile analysis of complexity；
6) the blob image block for meeting complexity threshold carries out feature extraction and compares with model, obtains water detection result.
As a further solution of the present invention: in step 1), outdoor scene on daytime ponding model training is specifically included: interception
Ponding region picture in network picture, establishes positive sample library；It randomly selects any non-ponding region picture and establishes negative example base；Just
The normalization of negative sample size；Positive negative sample gradient information and texture information are extracted as feature, combining classification device is trained
To outdoor scene on daytime ponding model.
As a further solution of the present invention: ponding region picture is parking stall ponding picture.
As a further solution of the present invention: in step 4), analyzing pixel color characteristics and saturation degree characteristic is specifically wrapped
It includes: RGB image in region is transformed into the space YCrCb and HSV space respectively, whether analyze each component value in given threshold range
It is interior.
As a further solution of the present invention: in step 5), profile analysis of complexity is specifically included: calculating candidate ponding
The elemental area and circumference pixel perimeter of blob image block, the ratio for analyzing elemental area and circumference pixel perimeter are
It is no to reach given threshold requirement, think to meet the requirement of profile complexity if reaching given threshold and requiring.
A kind of system that parking stall water detection is carried out based on image processing techniques, by being formed with lower module:
Ponding model training module: for carrying out outdoor scene on daytime ponding model training；
Fish eye lens demarcating module: for carrying out backsight fish eye lens calibration, internal reference matrix and distortion factor matrix are saved；
Image distortion correction module: for obtaining the video flowing of backsight fish eye lens acquisition, distortion correction is made to image；
Color characteristics analysis module: for setting detection zone, color notation conversion space is made to image in region, analyzes pixel
Point color characteristics and saturation degree retain the pixel for meeting given threshold；
Profile analysis of complexity module: for carrying out Morphological scale-space and connected component analysis, meet area threshold
Blob image block makees profile analysis of complexity；
Feature extraction and model comparison module: for the blob image block for meeting complexity threshold carry out feature extraction with
Model compares, and obtains water detection result.
As a further solution of the present invention: the outdoor scene ponding model training on daytime specifically includes: intercept network
Ponding region picture in picture establishes positive sample library；It randomly selects any non-ponding region picture and establishes negative example base；Positive and negative sample
The normalization of this size；Positive negative sample gradient information and texture information are extracted as feature, combining classification device is trained to obtain room
Outer scene on daytime ponding model.
As a further solution of the present invention: ponding region picture is parking stall ponding picture.
As a further solution of the present invention: the analysis pixel color characteristics and saturation degree characteristic specifically include: will
RGB image is transformed into the space YCrCb and HSV space respectively in region, whether within the set threshold range to analyze each component value.
As a further solution of the present invention: the profile analysis of complexity specifically includes: calculating candidate ponding blob figure
As the elemental area and circumference pixel perimeter of block, whether the ratio for analyzing elemental area and circumference pixel perimeter reaches
Given threshold requirement is thought to meet the requirement of profile complexity if reaching given threshold and requiring.
Compared with prior art, the beneficial effects of the present invention are:
A kind of method and system carrying out parking stall water detection based on image processing techniques proposed by the present invention, can be direct
Parking stall internal water accumulation situation is analyzed using image sequence characteristic, detection of obstacles in parking stall during further perfect intelligent parking
Problem provides auxiliary information to be unmanned.The present invention establishes the vehicle on intelligent automobile concept fisheye camera Information base
Position water detection technology, can quickly and efficiently identify parking stall ponding, more be able to satisfy the intelligent demand of parking of user.
Detailed description of the invention
Fig. 1 is flow chart of the method for the present invention；
Fig. 2 is system block diagram of the invention.
Specific embodiment
Below in conjunction with the embodiment of the present invention, technical scheme in the embodiment of the invention is clearly and completely described,
Obviously, described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.Based in the present invention
Embodiment, every other embodiment obtained by those of ordinary skill in the art without making creative efforts, all
Belong to the scope of protection of the invention.
Embodiment 1
In the embodiment of the present invention, as shown in Figure 1, for a kind of method for carrying out parking stall water detection based on image processing techniques
Flow chart, which comprises
S101. scene ponding model training on daytime outside room；
S102. backsight fish eye lens is demarcated, and saves internal reference matrix and distortion factor matrix；
S103. the video flowing for obtaining the acquisition of backsight fish eye lens, makees distortion correction to image；
S104. detection zone is set, color notation conversion space is made to image in region, analyzes pixel color characteristics and saturation
Degree retains the pixel for meeting given threshold；
S105. Morphological scale-space and connected component analysis, the blob image block for meeting area threshold make profile complexity point
Analysis；
S106. the blob image block for meeting complexity threshold carries out feature extraction and compares with model, obtains water detection knot
Fruit.
Ponding region picture in intercept network picture, especially parking stall ponding picture, establish positive sample library；It randomly selects and appoints
Non- ponding region picture of anticipating establishes negative example base；Positive and negative sample-size normalization；
Positive negative sample gradient information is extracted using histogramming algorithms such as EOH or HOG, is extracted using Gabor or LBP scheduling algorithm
Texture information merges gradient and texture as the sample characteristics；Positive negative sample is carried out using classifiers such as SVM or Adaboost
Training, obtains outdoor scene on daytime ponding model；
Black and white gridiron pattern is made, it is highly 10 lattice, grid length and width are 6cm that width, which is 14 lattice,；
Camera calibration is carried out using the picture of the backsight fish-eye camera shooting above different angle of 10 width, obtains internal reference and abnormal
Variable coefficient matrix simultaneously saves；
It parks after starting, obtains the video flowing of backsight fish eye lens acquisition in real time, calibrating parameters is called to carry out pattern distortion
Correction；
Detection zone is set, RGB image in region is transformed into the space YCrCb and HSV space respectively, analysis Cr, Cb,
Whether Cr-Cb and S etc. divides magnitude within the set threshold range to retain if within the set threshold range, otherwise set pixel value
It is zero.
The elemental area and circumference pixel perimeter for calculating candidate ponding blob image block, analyze elemental area and periphery
Whether the ratio of contour pixel perimeter is greater than given threshold, retains if reaching given threshold and requiring, is otherwise set as pixel value
Zero.
Feature extraction and and mould are carried out to the blob image block for meeting complexity threshold also with gradient and texture operator
Type feature compares, and obtains water detection result.
As shown in Fig. 2, the embodiment of the present invention 1 also proposed and a kind of carry out parking stall water detection based on image processing techniques
System, the system are achieved through the following technical solutions:
A kind of system that parking stall water detection is carried out based on image processing techniques, the system comprises:
Ponding model training module: for carrying out outdoor scene on daytime ponding model training；
Fish eye lens demarcating module: for carrying out backsight fish eye lens calibration, internal reference matrix and distortion factor matrix are saved；
Image distortion correction module: for obtaining the video flowing of backsight fish eye lens acquisition, distortion correction is made to image；
Color characteristics analysis module: for setting detection zone, color notation conversion space is made to image in region, analyzes pixel
Point color characteristics and saturation degree retain the pixel for meeting given threshold；
Profile analysis of complexity module: for carrying out Morphological scale-space and connected component analysis, meet area threshold
Blob image block makees profile analysis of complexity；
Feature extraction and model comparison module: for the blob image block for meeting complexity threshold carry out feature extraction with
Model compares, and obtains water detection result.
A kind of method and system that parking stall water detection is carried out based on image processing techniques that the embodiment of the present invention proposes, can
Directly to utilize image sequence characteristic to analyze parking stall internal water accumulation situation, obstacle in parking stall during further perfect intelligent parking
Analyte detection problem provides auxiliary information to be unmanned.Therefore, the embodiment of the present invention is established in intelligent automobile concept flake phase
Parking stall water detection technology on the basis of machine information, is compared compared with method, can quickly and efficiently identify parking stall ponding, more
It is able to satisfy the intelligent demand of parking of user.
It is obvious to a person skilled in the art that invention is not limited to the details of the above exemplary embodiments, Er Qie
In the case where without departing substantially from spirit or essential attributes of the invention, the present invention can be realized in other specific forms.Therefore, no matter
From the point of view of which point, the present embodiments are to be considered as illustrative and not restrictive, and the scope of the present invention is by appended power
Benefit requires rather than above description limits, it is intended that all by what is fallen within the meaning and scope of the equivalent elements of the claims
Variation is included within the present invention.
In addition, it should be understood that although this specification is described in terms of embodiments, but not each embodiment is only wrapped
Containing an independent technical solution, this description of the specification is merely for the sake of clarity, and those skilled in the art should
It considers the specification as a whole, the technical solutions in the various embodiments may also be suitably combined, forms those skilled in the art
The other embodiments being understood that.
Claims (2)
1. a kind of method for carrying out parking stall water detection based on image processing techniques, which comprises the following steps:
1) outdoor scene ponding model training on daytime；Outdoor scene on daytime ponding model training specifically includes: intercept network picture
Middle parking stall ponding picture, establishes positive sample library；It randomly selects any non-parking stall ponding picture and establishes negative example base；Positive negative sample ruler
Very little normalization；Positive negative sample gradient information and texture information are extracted as feature, combining classification device is trained to obtain outdoor white
Its scene ponding model；
2) backsight fish eye lens is demarcated, and saves internal reference matrix and distortion factor matrix；
3) video flowing for obtaining the acquisition of backsight fish eye lens, makees distortion correction to image；
4) detection zone is set, color notation conversion space is made to image in region, analyzes pixel color characteristics and saturation degree, is retained
Meet the pixel of given threshold；Analysis pixel color characteristics and saturation degree characteristic specifically include: by RGB image in region point
It is not transformed into the space YCrCb and HSV space, whether within the set threshold range to analyze each component value；
5) Morphological scale-space and connected component analysis, the blob image block for meeting area threshold make profile analysis of complexity；Profile
Analysis of complexity specifically includes: calculating the elemental area and circumference pixel perimeter of candidate ponding blob image block, analyzes picture
Whether vegetarian noodles product and the ratio of circumference pixel perimeter reach given threshold requirement, think full if reaching given threshold and requiring
Sufficient profile complexity requirement；
6) the blob image block for meeting complexity threshold carries out feature extraction and compares with model, obtains water detection result.
2. a kind of system for carrying out parking stall water detection based on image processing techniques, which is characterized in that by being formed with lower module:
Ponding model training module: for carrying out outdoor scene on daytime ponding model training；Outdoor scene on daytime ponding model instruction
White silk specifically includes: ponding picture in parking stall in intercept network picture establishes positive sample library；Randomly select any non-parking stall ponding picture
Establish negative example base；Positive and negative sample-size normalization；Positive negative sample gradient information and texture information are extracted as feature, in conjunction with point
Class device is trained to obtain outdoor scene on daytime ponding model；
Fish eye lens demarcating module: for carrying out backsight fish eye lens calibration, internal reference matrix and distortion factor matrix are saved；
Image distortion correction module: for obtaining the video flowing of backsight fish eye lens acquisition, distortion correction is made to image；
Color characteristics analysis module: for setting detection zone, color notation conversion space is made to image in region, analyzes pixel face
Color characteristic and saturation degree retain the pixel for meeting given threshold；Analysis pixel color characteristics and saturation degree characteristic are specifically wrapped
It includes: RGB image in region is transformed into the space YCrCb and HSV space respectively, whether analyze each component value in given threshold range
It is interior；
Profile analysis of complexity module: for carrying out Morphological scale-space and connected component analysis, meet the blob figure of area threshold
As block makees profile analysis of complexity；Profile analysis of complexity specifically includes: calculating the elemental area of candidate ponding blob image block
And whether circumference pixel perimeter, the ratio for analyzing elemental area and circumference pixel perimeter reach given threshold requirement,
Think to meet the requirement of profile complexity if reaching given threshold and requiring；
Feature extraction and model comparison module: for carrying out feature extraction and model to the blob image block for meeting complexity threshold
It compares, obtains water detection result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610129872.7A CN105809131B (en) | 2016-03-08 | 2016-03-08 | A kind of method and system carrying out parking stall water detection based on image processing techniques |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610129872.7A CN105809131B (en) | 2016-03-08 | 2016-03-08 | A kind of method and system carrying out parking stall water detection based on image processing techniques |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105809131A CN105809131A (en) | 2016-07-27 |
| CN105809131B true CN105809131B (en) | 2019-10-15 |
Family
ID=56467836
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610129872.7A Active CN105809131B (en) | 2016-03-08 | 2016-03-08 | A kind of method and system carrying out parking stall water detection based on image processing techniques |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105809131B (en) |
Families Citing this family (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107862679B (en) * | 2017-10-20 | 2020-12-18 | 歌尔光学科技有限公司 | Method and device for determining image detection area |
| CN107818685A (en) * | 2017-10-25 | 2018-03-20 | 司法部司法鉴定科学技术研究所 | A kind of method that state of motion of vehicle is obtained based on Vehicular video |
| CN107909070A (en) * | 2017-11-24 | 2018-04-13 | 天津英田视讯科技有限公司 | A kind of method of road water detection |
| CN111219922A (en) * | 2018-11-23 | 2020-06-02 | 宁波泽锦电器科技有限公司 | Multifunctional horizontal refrigerator |
| CN111027461B (en) * | 2019-12-06 | 2022-04-29 | 长安大学 | Vehicle track prediction method based on multi-dimensional single-step LSTM network |
| CN112193240B (en) * | 2020-09-28 | 2022-02-01 | 惠州华阳通用电子有限公司 | Parking method based on water accumulation information |
| CN112172798B (en) * | 2020-09-28 | 2022-02-01 | 惠州华阳通用电子有限公司 | Parking method based on water accumulation environment and storage medium |
| CN113044023B (en) * | 2020-09-30 | 2022-04-01 | 惠州华阳通用电子有限公司 | Parking space ponding identification method and device |
| CN112842184B (en) * | 2021-02-25 | 2022-11-29 | 深圳银星智能集团股份有限公司 | Cleaning method and cleaning robot |
| CN114820468A (en) * | 2022-04-06 | 2022-07-29 | 上海擎测机电工程技术有限公司 | Accumulated water detection method based on color-changing paper and image recognition |
| CN114758139B (en) * | 2022-06-16 | 2022-10-21 | 成都鹏业软件股份有限公司 | Method for detecting accumulated water in foundation pit |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| FR2901218B1 (en) * | 2006-05-22 | 2009-02-13 | Valeo Vision Sa | METHOD FOR DETECTING RAIN ON A SPRINKLER |
| US9460353B2 (en) * | 2010-09-16 | 2016-10-04 | California Institute Of Technology | Systems and methods for automated water detection using visible sensors |
| CN102901489B (en) * | 2011-07-25 | 2016-09-07 | 烟台惠通网络技术有限公司 | Surface gathered water accumulated ice detection method and device |
| CN102509098B (en) * | 2011-10-08 | 2013-05-15 | 天津大学 | Fisheye image vehicle identification method |
| CN103116743B (en) * | 2013-02-01 | 2016-08-10 | 浙江捷尚视觉科技股份有限公司 | A kind of railway obstacle detection method based on on-line study |
-
2016
- 2016-03-08 CN CN201610129872.7A patent/CN105809131B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN105809131A (en) | 2016-07-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105809131B (en) | A kind of method and system carrying out parking stall water detection based on image processing techniques | |
| Setjo et al. | Thermal image human detection using Haar-cascade classifier | |
| Shaik et al. | Comparative study of skin color detection and segmentation in HSV and YCbCr color space | |
| Ajmal et al. | A comparison of RGB and HSV colour spaces for visual attention models | |
| US20140240477A1 (en) | Multi-spectral imaging system for shadow detection and attenuation | |
| CN109271921A (en) | A kind of intelligent identification Method and system of multispectral imaging | |
| US10140521B2 (en) | Method, system and apparatus for processing an image | |
| CN104318262A (en) | Method and system for replacing skin through human face photos | |
| WO2023155324A1 (en) | Image fusion method and apparatus, device and storage medium | |
| US20160092726A1 (en) | Using gestures to train hand detection in ego-centric video | |
| US9418426B1 (en) | Model-less background estimation for foreground detection in video sequences | |
| WO2020223963A1 (en) | Computer-implemented method of detecting foreign object on background object in image, apparatus for detecting foreign object on background object in image, and computer-program product | |
| Qian et al. | Revisiting gray pixel for statistical illumination estimation | |
| Amjad et al. | Multiple face detection algorithm using colour skin modelling | |
| CN112101260B (en) | Method, device, equipment and storage medium for identifying safety belt of operator | |
| US20110158511A1 (en) | Method and apparatus for filtering red and/or golden eye artifacts | |
| CN113128308B (en) | Pedestrian detection method, device, equipment and medium in port scene | |
| CN112926367B (en) | Living body detection equipment and method | |
| Low et al. | Experimental study on multiple face detection with depth and skin color | |
| Lin et al. | Face detection based on skin color segmentation and SVM classification | |
| Mahmoodi et al. | Boosting performance of face detection by using an efficient skin segmentation algorithm | |
| Wang et al. | Deep learning-based human activity analysis for aerial images | |
| CN116188859A (en) | Tea disease unmanned aerial vehicle remote sensing monitoring method based on superdivision and detection network | |
| CN108256401B (en) | Method and device for obtaining target attribute feature semantics | |
| Wu et al. | Real-time 2D hands detection and tracking for sign language recognition |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20200610 Address after: 200333 room 124, No.3, Lane 129, Lane 822, Zhennan Road, Putuo District, Shanghai Patentee after: Shanghai yingma Information Technology Co.,Ltd. Address before: 315700 Ningbo, Xiangshan Province Economic Development Zone, Park Road, No. 98 Patentee before: NINGBO YULAN INFORMATION TECHNOLOGY Co.,Ltd. |
|
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20220818 Address after: Room 1606, Youte Headquarters Building, No. 88, Xingye Road, Xiangzhou District, Zhuhai City, Guangdong Province, 519000 Patentee after: Zhuhai Daxuan Information Technology Co., Ltd. Address before: Room 124, No. 3, Lane 129, Lane 129, Lane 822, Zhennan Road, Putuo District, Shanghai 200333 Patentee before: Shanghai yingma Information Technology Co.,Ltd. |
|
| TR01 | Transfer of patent right |