CN103985134B - It is a kind of to look around the detection method for demarcating synthetic effect - Google Patents
It is a kind of to look around the detection method for demarcating synthetic effect Download PDFInfo
- Publication number
- CN103985134B CN103985134B CN201410245590.4A CN201410245590A CN103985134B CN 103985134 B CN103985134 B CN 103985134B CN 201410245590 A CN201410245590 A CN 201410245590A CN 103985134 B CN103985134 B CN 103985134B
- Authority
- CN
- China
- Prior art keywords
- detection
- vehicle
- line
- viewing system
- image
- 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.)
- Expired - Fee Related
Links
Abstract
It is a kind of to look around the detection method for demarcating synthetic effect, by building hough transform region, put detection block and test bar, by adjust detection vehicle location, glacing flatness detection, splicing accuracy detection, can identification detection and blind area detection reach detection to vehicle-mounted viewing system camera parameter calibration accuracy, it can be ensured that the accuracy of demarcation.
Description
Technical field
The present invention relates to vehicle-mounted camera parameter calibration detection field, and in particular to a kind of to look around the inspection for demarcating synthetic effect
Survey method.
Background technology
Vehicle-mounted viewing system is made up of the multiple stage wide angle cameras being assemblied on car body all around, and every video camera covers
Lid is certain interval, and image generates top view through viewpoint change, and carries out splicing and obtain vehicle-mounted looking around aerial view.In order that above-mentioned
The image absorbed by the multiple stage wide angle cameras being assemblied on car body can generate the seamless spelling not distorted through viewpoint change
Panoramic view is looked down in the synthesis for connecing, and inner parameter and the external parameter of the above-mentioned wide angle cameras being assemblied on car body must be accurate
Demarcate.Whether the inner parameter of video camera accurately directly affects the distortion correction effect of composite diagram.The external parameter of video camera is
It is no can accurately affect composite diagram in each camera image junction whether misplace, look down panoramic view in also affecting final composite diagram
It is whether correct with the relative position relation of car body.
In the system, the part of most critical is the demarcation to camera inner parameter and external parameter, and synthesis process
Image algorithm.Prior art and product adopt a variety of algorithms, cause in the deformation extent for getting a bird's eye view picture for ultimately forming,
Degreeof tortuosity, the laminating degree of the splicing of each camera picture, and fog-level of process fusion picture etc. have very big difference.
Meanwhile, not to looking around the method for inspection and foundation of demarcating composite result in prior art and product.How to set
Meter demarcates the detection method of effect, to guarantee the accuracy of camera parameter calibration, becomes the weight of vehicle-mounted viewing system research
Want direction.
The content of the invention
Goal of the invention：In order to overcome the deficiencies in the prior art, the present invention to provide a kind of detection looked around and demarcate synthetic effect
Method, by setting detection zone on ground, by the position adjustment of the area image, glacing flatness detection, accuracy detection, distinguishable
Knowledge and magnanimity are detected and blind area is detected, it is ensured that the accuracy of camera parameter calibration.
Technical scheme：A kind of vehicle-mounted viewing system demarcates the detection method of synthetic effect, including host computer；It is characterized in that
Including an actual place detection zone, the actual place detection zone is rectangle, position on the basis of the center of the rectangle；It is flat
Row has some detection lines in long side and the broadside of rectangular area, respectively spaced set；Square has been evenly arranged in detection line
Detection block；Including test bar；
The method is comprised the following steps：
1) adjustment detection vehicle location：The actual place detection zone of detection vehicle viewing system output is obtained by host computer
Area image, the such as image are mismatched with detection zone image, are chosen adjusting point, are converted by plane Affine, obtain detection zone
The remedial frames that domain is matched with actual place detection zone；
2) glacing flatness detection：The detection line includes horizontal detection line and longitudinal detection line, calculates detection line on image
, respectively to standard lateral and the mean value of standard longitudinal direction detection line vertical range, computing formula is as follows for each pixel：
Wherein,
NhFor the summation of all horizontal detection line pixel counts；
NvFor the summation of all longitudinal detection line pixel counts；
HiFor preferable i-th x wire,For i-th x wire for detecting；
VjFor preferable j-th strip vertical line,For the j-th strip vertical line for detecting；
Dist () is the vertical range summation function put in above-mentioned two lines；
3) splice accuracy detection：In viewing system, boundary line and the detection line phase of every camera FOV intersection
Hand over, judge whether the detection line on image has bifurcated and fracture by the staggered pixels at the joining or dislocation；
4) can identification detection：By datum level place detection block, with detection block detect viewing system can identification；
Detection block in observation imaging whether there is, if fuzzy, deformation, if edge clear is visible.
5) blind area detection：The test bar is placed perpendicular to ground, one end patch ground；Test bar and the vehicle body for detecting vehicle
Fixed range is kept to move some weeks, the distance for controlling test bar each week with vehicle body is incremented by；Check whether output image occurs
Blind area or disjunction phenomenon；
The detection block length of side is 5cm-10cm.
The a diameter of 5cm of test bar, length is 1.5m.
Step 5) in, the increasing lengths of test bar and vehicle body distance according to 10cm, 30cm, 50cm, 100cm.Such size
Inspection block and check rod be easy to operate and recognize.
Beneficial effect：
1st, by a series of detection, ensure that the accuracy of vehicle-mounted viewing system camera parameter calibration；
2nd, the reducing degree looked around to actual vehicle and surrounding enviroment, including visual can be detected, whether ratio is accurate
Really, vehicle grazing can not occur because of proportional imbalance.
3rd, the precision for looking around picture is can interpolate that, lines distortion or fuzzy on picture is looked around, be can confirm that such ring
Depending on putting in actual use, it is impossible to which effectively auxiliary human pilot operates vehicle.
4th, the size and location of blind area is can interpolate that, more than a certain degree of blind area scope, you can confirm the viewing system
It is unqualified.
Description of the drawings
Fig. 1 actual place detection zone schematic diagrames
Before Fig. 2 adjustment vehicle detection area images
After Fig. 3 adjustment vehicle detection area images
Fig. 4 glacing flatness detects schematic diagrams
Fig. 5 boundary line schematic diagrames
Fig. 6 vertical directions imaging dislocation schematic diagram
Fig. 7 puts schematic diagram for detection block
Fig. 8 test bar mobile route schematic diagrames
Specific embodiment
Below in conjunction with the accompanying drawings the present invention is done and is further explained.
A kind of vehicle-mounted viewing system demarcates the detection method of synthetic effect, including host computer；In this example, imaging is referred to
The image that the picture that the camera of vehicle-mounted viewing system shoots is shown after passing host computer back.
As shown in figure 1, the method includes an actual place detection zone, the place is set up on smooth ground, real
Border place detection zone is rectangle, is to ensure that most of dilly can enter inspection, the size of the rectangular area of this example
It is set to 3m × 6m.Position on the basis of the center of the rectangle is set, for positioning vehicle.Detect other kinds of large-scale if desired
Vehicle, the method for arranging place are identical, proportionally adjust accordingly.
As shown in figure 1, actual place detection zone is provided with detection line, long side of these detection lines parallel to rectangular area
And broadside；Including horizontal detection line and longitudinal detection line, width of the horizontal detection line parallel to rectangle in figure, longitudinal detection line are put down
Row is in the height of rectangle；Wherein, there are two detection lines intersecting through reference position.
The method is comprised the following steps：
Step 1) adjustment detection vehicle location：The actual place inspection of detection vehicle viewing system output is obtained by host computer
Area image is surveyed, as shown in Fig. 2 because of storing cycle not just, the image and detection zone image captured by vehicle-mounted viewing system
Mismatch, need image correction, the detection zone image of the rectangular area allowed in captured image and host computer Plays
Essentially coincide.Artificial operation is now needed, adjusting point is chosen, this example selects two tops in rectangular area upper left and upper right in Fig. 2
Point, host computer are converted automatically by plane Affine, are modified to position and the angle of picture, so as to obtain detection zone and actual field
The remedial frames that ground detection zone matches, the remedial frames are as shown in Figure 3.
2) glacing flatness detection：For detecting the precision of camera inner parameter calculating.Judge each bar detection line glacing flatness
Whether set test stone is met.Concrete grammar is each pixel for calculating detection line on image respectively to standard lateral and mark
The mean value of quasi- longitudinal direction detection line vertical range, computing formula are as follows：
Wherein,
NhFor the summation of all horizontal detection line pixel counts；
NvFor the summation of all longitudinal detection line pixel counts；
HiFor preferable i-th x wire,For i-th x wire for detecting；
VjFor preferable j-th strip vertical line,For the j-th strip vertical line for detecting；
Dist () is the vertical range summation function put in above-mentioned two lines；
As shown in figure 4, line straight in figure is normal line, above normal line, not straight line is exported to be detected camera
Detection line.By calculating, when the mean value of vertical range exceeds given area, that is, judge that glacing flatness detection is unqualified.
3) splice accuracy detection：In viewing system, as shown in figure 5, four oblique lines in figure are every camera ken model
The boundary line of intersection is enclosed, the boundary line is intersected with detection line, the point intersected with horizontal item detection line can aid in detection level side
To splicing precision.The point intersected with longitudinal detection line, can aid in detecting the splicing precision of vertical direction.As shown in fig. 6, logical
Whether the detection line that the staggered pixels or dislocation crossed at the joining are judged on image has bifurcated and fracture；The d1 represented in Fig. 6
For the width of dislocation, d2 is normal live width.
4) can identification detection：As shown in fig. 7, by detection block detect viewing system can identification；Detection block is for just
Square, the length of side is set in 5cm-10cm as needed.Uniformly put along detection line.Whether the detection block in observation imaging deposits
In, if fuzzy, deformation, if edge clear is visible etc..
5) blind area detection：The test bar is placed perpendicular to ground, one end patch ground；The material of test bar may be selected PVC
The common used materials such as tubing, can recycle.The diameter of test bar is about 5cm, and length is about 1.5m.
Test bar keeps fixed range movement mobile one week with the vehicle body of detection vehicle, forms a detection circle, can choose
The each movement in 10cm, 30cm, 50cm, 100cm source one week, mobile route is as shown in Figure 8.Check whether output image blind area occurs
Or disjunction phenomenon.
In a practical situation, in the case of conditions permit, species can be increased and is separately verified.Check rod and vehicle body distance
Less value can be used, and with multiple different lengths checking, the accuracy of blind area detection is improved by repeated detection.
The above is only the preferred embodiment of the present invention, it is noted that for the ordinary skill people of the art
For member, under the premise without departing from the principles of the invention, some improvements and modifications can also be made, these improvements and modifications also should
It is considered as protection scope of the present invention.
Claims (4)
1. a kind of vehicle-mounted viewing system demarcates the detection method of synthetic effect, including host computer；It is characterized in that including a reality
Border place detection zone, the actual place detection zone are rectangle, position on the basis of the center of the rectangle, also, the place
Set up on smooth ground, actual place detection zone is rectangle, and to ensure that most of vehicle can enter inspection；It is parallel
In long side and the broadside of rectangular area, spaced set has some detection lines respectively；It has been evenly arranged in detection line foursquare
Detection block, wherein, detection block is put evenly along detection line；Including test bar；
The method is comprised the following steps：
1) adjustment detection vehicle location：The actual place detection zone figure of detection vehicle viewing system output is obtained by host computer
Picture, the such as image are mismatched with detection zone image, are chosen adjusting point, are converted by plane Affine, obtain detection zone with
The remedial frames that actual place detection zone matches；
2) glacing flatness detection：The detection line includes horizontal detection line and longitudinal detection line, calculates detection line on image each
, respectively to standard lateral and the mean value of standard longitudinal direction detection line vertical range, computing formula is as follows for pixel：
Wherein,
NhFor the summation of all horizontal detection line pixel counts；
NvFor the summation of all longitudinal detection line pixel counts；
HiFor preferable i-th x wire,For i-th x wire for detecting；
VjFor preferable j-th strip vertical line,For the j-th strip vertical line for detecting；
Dist () is the vertical range summation function put in two lines；Wherein, when the mean value of vertical range exceeds given area,
Judge that glacing flatness detection is unqualified；
3) splice accuracy detection：In viewing system, the boundary line of every camera FOV intersection is intersected with detection line, is led to
Whether the detection line that the staggered pixels or dislocation crossed at the joining are judged on image has bifurcated and fracture；
4) can identification detection：By datum level place detection block, with detection block detect viewing system can identification；Observation
Detection block in imaging whether there is, if fuzzy, deformation, if edge clear is visible；
5) blind area detection：The test bar is placed perpendicular to ground, one end patch ground；Test bar is kept with the vehicle body of detection vehicle
Fixed range moves some weeks, and the distance for controlling test bar each week with vehicle body is incremented by；Check whether output image blind area occurs
Or disjunction phenomenon.
2. a kind of vehicle-mounted viewing system demarcates the detection method of synthetic effect as claimed in claim 1, it is characterised in that the inspection
It is 5cm-10cm to survey the block length of side.
3. a kind of vehicle-mounted viewing system demarcates the detection method of synthetic effect as claimed in claim 1, it is characterised in that the inspection
The a diameter of 5cm of rod is surveyed, length is 1.5m.
4. a kind of vehicle-mounted viewing system demarcates the detection method of synthetic effect as claimed in claim 1, it is characterised in that the step
It is rapid 5) in, the increasing lengths of test bar and vehicle body distance according to 10cm, 30cm, 50cm, 100cm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410245590.4A CN103985134B (en) | 2014-06-04 | 2014-06-04 | It is a kind of to look around the detection method for demarcating synthetic effect |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410245590.4A CN103985134B (en) | 2014-06-04 | 2014-06-04 | It is a kind of to look around the detection method for demarcating synthetic effect |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103985134A CN103985134A (en) | 2014-08-13 |
| CN103985134B true CN103985134B (en) | 2017-04-05 |
Family
ID=51277092
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410245590.4A Expired - Fee Related CN103985134B (en) | 2014-06-04 | 2014-06-04 | It is a kind of to look around the detection method for demarcating synthetic effect |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103985134B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9902322B2 (en) | 2015-10-30 | 2018-02-27 | Bendix Commercial Vehicle Systems Llc | Filling in surround view areas blocked by mirrors or other vehicle parts |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6236748B1 (en) * | 1994-08-02 | 2001-05-22 | Canon Kabushiki Kaisha | Compound eye image pickup device utilizing plural image sensors and plural lenses |
| CN101079151A (en) * | 2006-10-13 | 2007-11-28 | 浙江师范大学 | 360 degree around panorama generation method based on serial static image |
| CN101976460A (en) * | 2010-10-18 | 2011-02-16 | 胡振程 | Generating method of virtual view image of surveying system of vehicular multi-lens camera |
-
2014
- 2014-06-04 CN CN201410245590.4A patent/CN103985134B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6236748B1 (en) * | 1994-08-02 | 2001-05-22 | Canon Kabushiki Kaisha | Compound eye image pickup device utilizing plural image sensors and plural lenses |
| CN101079151A (en) * | 2006-10-13 | 2007-11-28 | 浙江师范大学 | 360 degree around panorama generation method based on serial static image |
| CN101976460A (en) * | 2010-10-18 | 2011-02-16 | 胡振程 | Generating method of virtual view image of surveying system of vehicular multi-lens camera |
Non-Patent Citations (2)
| Title |
|---|
| Robust metric calibration of non-linear camera lens distortion;Ricolfe-Viala Carlos等;《Pattern Recognition》;20101231;第43卷(第4期);第1688－1699页 * |
| 面向坦克的全景环视系统关键技术研究;徐杰;《中国优秀硕士学位论文全文数据库 信息科技辑》;20140315(第03期);第2页第1.2节；第4页第1.4.1节1－6行，第28页第3.5.1节图3.6（a）（b）；第11页第2.2.3.1节第1－9行；第9页第1－23行 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103985134A (en) | 2014-08-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105674880B (en) | Contact net geometric parameter measurement method and system based on binocular principle | |
| CN103411553B (en) | The quick calibrating method of multi-linear structured light vision sensors | |
| US11482013B2 (en) | Object tracking method, object tracking apparatus, vehicle having the same, and computer-program product | |
| US20130147948A1 (en) | Image processing apparatus and imaging apparatus using the same | |
| CN106485751B (en) | Unmanned aerial vehicle photographic imaging and data processing method and system applied to foundation pile detection | |
| CN105405126B (en) | A kind of multiple dimensioned vacant lot parameter automatic calibration method based on single camera vision system | |
| CN106645205A (en) | Unmanned aerial vehicle bridge bottom surface crack detection method and system | |
| CN101697105B (en) | Camera type touch detection positioning method and camera type touch detection system | |
| CN103186892B (en) | Aerial Images is utilized to generate the method and system of equal proportion outdoor scene scene photo | |
| CN104484887B (en) | External parameters calibration method when video camera is used in combination with scanning laser range finder | |
| CN109685858A (en) | A kind of monocular cam online calibration method | |
| DE102011100628B4 (en) | Method and device for determining at least one camera parameter | |
| CN103913149B (en) | A kind of binocular range-measurement system and distance-finding method thereof based on STM32 single-chip microcomputer | |
| CN103852060A (en) | Visible light image distance measuring method based on monocular vision | |
| CN103729837A (en) | Rapid calibration method of single road condition video camera | |
| CN109828267A (en) | The Intelligent Mobile Robot detection of obstacles and distance measuring method of Case-based Reasoning segmentation and depth camera | |
| CN113223075A (en) | Ship height measuring system and method based on binocular camera | |
| CN104853154B (en) | A kind of extraction of crag deformation information and alarm method based on moving image | |
| CN102446355B (en) | Method for detecting target protruding from plane based on double viewing fields without calibration | |
| CN108469254A (en) | A kind of more visual measuring system overall calibration methods of big visual field being suitable for looking up and overlooking pose | |
| CN101776437B (en) | Calibration technology for vision sub-pixel of embedded type machine with optical path adjustment | |
| CN112836634B (en) | Multi-sensor information fusion gate anti-trailing method, device, equipment and medium | |
| KR101771492B1 (en) | Method and system for mapping using UAV and multi-sensor | |
| CN101494772A (en) | Method and apparatus for detecting image | |
| CN107564070A (en) | The ranging of large scene binocular and its bearing calibration in the monitoring of overhead power transmission channel image |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C41 | Transfer of patent application or patent right or utility model | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20170106 Address after: 214000 Jiangsu City, Wuxi New District Longshan Road, No. 4 (Wang Zhuang science and Technology Venture Center) C building 203-204 Applicant after: Wuxi Wissen Intelligent Sensing Technology Co., Ltd. Address before: 214028 Wuxi Province, Jiangsu New District Longshan Road, No. 4 (Wang Zhuang science and Technology Venture Center) C building, floor 2 Applicant before: Wuxi Guan Zhi vision Science and Technology Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CP01 | Change in the name or title of a patent holder | ||
| CP01 | Change in the name or title of a patent holder |
Address after: 214000 C 203-204, Longshan Road, Wuxi New District, Jiangsu, China (203-204) Patentee after: Wuxi wisdom Sensor Technology Co., Ltd. Address before: 214000 C 203-204, Longshan Road, Wuxi New District, Jiangsu, China (203-204) Patentee before: Wuxi Wissen Intelligent Sensing Technology Co., Ltd. |
|
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20170405 Termination date: 20190604 |