CN106504287A - Monocular vision object space alignment system based on template - Google Patents
Monocular vision object space alignment system based on template Download PDFInfo
- Publication number
- CN106504287A CN106504287A CN201610910942.2A CN201610910942A CN106504287A CN 106504287 A CN106504287 A CN 106504287A CN 201610910942 A CN201610910942 A CN 201610910942A CN 106504287 A CN106504287 A CN 106504287A
- Authority
- CN
- China
- Prior art keywords
- central point
- summit
- template
- calibrating
- 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.)
- Granted
Links
Landscapes
- Image Analysis (AREA)
Abstract
Based on the monocular vision object space alignment system of template, belong to space orientation technique field, low for solving the problems, such as target positioning failure, positional accuracy, technical essential is:Including:Calibrating template, for calibrating camera, is used for detecting calibrating template image space two-dimensional position;Coordinate mapping module, calibrating template image space two-dimensional coordinate system is mapped to three-dimensional coordinate system, and derivation module, according to calibrating template in the plane of delineation projection derive its in three dimensions with respect to photographic head range-azimuth with to monocular vision target designation.Effect is:Moving target is replaced using template, template is detected, oriented and is found range, computation complexity is low, can fast and effectively position moving target.
Description
Technical field
The invention belongs to space orientation technique field, specifically a kind of monocular vision object space based on template is calmly
Position system.
Technical background
Being accurately positioned for target is played a very important role with analysis for the identification of target and the understanding of image, multiple
Target in miscellaneous background is positioned at the fields such as military affairs, industrial monitoring, traffic control and management important application.For target is determined
The research of position aspect, JorgeLob etc. propose the three-dimensional rebuilding method for combining Inertia information and vision, using inertial sensor with
Binocular vision combines to recover the three-dimensional parameter of ground level and line segment normal thereto;Chinese Academy of Sciences's Shenyang automation research Hao Ying
Bright grade presets some artificial targets in target, the three-dimensional information of environment is obtained according to binocular stereo vision, real
When calculate the position relationship of mobile robot relative flag thing;Stone is pure as jade to Double-visual angle moving target three-dimensional space in video monitoring
Between location technology studied, extracted based on SURF algorithm
Coordinate system realizes that target is positioned.
Monocular vision carries out a kind of sterically defined location technology by monocular-camera.Method not pregnancy ceased new at present
Raw, with multi-disciplinary cross reference, more promote monocular vision object localization method constantly to develop.Existing object localization method
Mainly target designation model is set up using visual signature, Target space position is determined by target projection positioning.Color, texture,
The visual signatures such as edge, light stream are easy to extract, but easily receive such environmental effects, and feature stability is bad, cause target positioning to be lost
Effect, positional accuracy are low.Although wavelet character, local feature, feature base etc. can strengthen the accuracy of positioning, feature extraction
Algorithm computation complexity is high, is unfavorable for that real-time target is positioned.Locating speed is fast, precision is high, the location algorithm of strong robustness is
The research emphasis of current monocular vision object localization method.
Content of the invention
When determining Target space position in order to solve existing object localization method by target projection positioning, its color, stricture of vagina
Though the visual signatures such as reason, edge, light stream are easy to extract, such environmental effects are easily received, feature stability is bad, cause target fixed
The low problem of position failure, positional accuracy.
Technical solution of the present invention is as follows:
A kind of monocular vision target designation system based on calibrating template, including:
Calibrating template, for calibrating camera, is used for detecting calibrating template image space two-dimensional position;
Calibrating template image space two-dimensional coordinate system is mapped to three-dimensional coordinate system by coordinate mapping module, and
Derivation module, the projection according to calibrating template in the plane of delineation derive which in three dimensions with respect to photographic head
Range-azimuth is with to monocular vision target designation.
Further, the calibrating template, adopts diagonal black squares square for template primitive form, with central point
With summit, the completeness of central point and summit is defined, opposite vertexes are encoded with central point.
Further,
The derivation module, including:
Parameter determination module:Calibrating template calibrating camera is made, initial alignment parameter is determined;
Image capture module:Calibrating template and target bind to be measured, carry out image acquisition;
Computing module:
The collection image angle point is searched for, complete and half complete central point and summit is extracted, calculating gathers its in image
In a pair of central points and summit Euclidean pixel distance;
Determine the collection image center that monocular vision is collected;
According to the central point and summit Euclidean pixel distance of the calibrating template of the collection image, and initial alignment parameter,
Calculate the Euclidean pixel distance of collection image center and camera center;
Calculate the Euclidean pixel distance of collection image center and calibrating template central point;
According to above-mentioned collection image center and camera center Euclidean pixel distance and
Collection image center and the Euclidean pixel distance of calibrating template central point,
Calculate distance, the azimuth at calibrating template center and camera center.
Further, the parameter determination module:
Calibrating template central point O points are positioned on camera lens group central axis, parallel to lens plane;
Calibrating template is taken pictures sampling to the translation of remote continuous horizontal by near, obtain each apart from upper uncalibrated image, respectively to often
Open uncalibrated image and detect all angle points, extract complete central point, half complete central point, complete summit and half complete summit, according to
The corresponding central point of vertex encoding coupling and summit, obtain central point and apex coordinate;
Calculate the Euclidean pixel distance d of 2 points of B, O in uncalibrated imageBO, the Euclidean pixel distance of 2 points of E, O in uncalibrated image
dEO:
Wherein, summit O (x0,y0), summit B (xB,yB), summit E (xE,yE);
dlRepresent Euclidean pixel distance between black patch summit and central point:
dl=E (dBO, dEO) (3)
Determine initial alignment parameter betai:
Wherein, DKOEuclidean space distance for uncalibrated image central point O and camera center point K.
Further, the computing module:
The central point of collection image is designated as P (x, y), according to central point and the summit Europe of the calibrating template of the collection image
Family name pixel distance dl, and initial alignment parameter betai, calculate the Euclidean space distance of collection image center P and camera center point K
DKP:
DKP=βi×dl(5)
Calibrating template central point O (x0,y0) with collection image center P (x, y) Euclidean pixel distance be dOP:
Wherein:(x0,y0) it is template central point O point coordinates in collection image.
The Euclidean space of calibrating template central point and camera center is apart from D and azimuth angle alpha:
Further, the central point of the calibrating template with the definition step of the completeness on summit is:
According to template form, the four-quadrant subregion on central point or summit is defined, four subregions are designated as I respectively1、I2、I3、I4,
The n pixel equidistantly chosen on any i directions by any central point or summit, its average isCorresponding central point or top
Point is in I1The pixel average of subregionFor:
In the same manner, the pixel average of four subregions of central point or summit is obtained with thisCentral point or top
Point completeness is defined as follows:
Complete central point or summit:
Half complete central point or summit:
Incomplete central point or summit:Do not exist;
Wherein, TH1Represent the similarity degree of diagonal black bars or diagonal white portion pixel in image, TH2Represent figure
The difference degree of black and white area pixel as in.
Further, the summit to calibrating template and comprising the concrete steps that central point is encoded:
Pixel average on 4 subregions of calibrating template central point O is remembered respectivelyEqual according to pixel
Value determines the coding of 4 subregions, and wherein black is with 0 coding, white with 1 coding;
Determine each subregion pixel average in summitArbitrary summit is encoded, wherein black uses 0
Coding, white are encoded with 1;
Pixel average according to four subregions of central point and summit determines that its color is encoded.
Beneficial effect:
Existing object locating system needs extra aiding sensors or distance-measuring equipment, needs to carry out coordinate system transformation or again
Modeling, computation complexity are high, it is impossible to provide accurate target component information for subsequent treatment in real time, cause target following, target
Identification, voice positioning etc. are processed and calculate failure, meanwhile, testing cost is too high, and multi-sensor information fusion also can reduce accuracy,
Higher to optimized algorithm requirement, it is unfavorable for that requirement of real-time height and the mini-plant of low-power consumption are adopted.
, in existing object space location technology, camera marking method is complicated for the present invention, it is impossible to realize in real time, accurately
The problem of positioning target, there is provided a kind of monocular vision object space alignment system based on calibrating template (afterwards referred to as in template
Also be calibrating template), the projection according to template in the plane of delineation derive its in three dimensions the distance of relative photographic head and
Orientation, when template close to be positioned over before moving target when, equally determined moving target can be positioned, be aided in this
Follow-up video image motion target following, moving target positioning, mike acoustic source array positioning and pedestrian target positioning etc.
Research.The system is to gather front end to can achieve the initialization positioning work to moving target merely with monocular cam/video camera
Make, save the coordinate system transformation step that is brought using the auxiliary positioning means such as laser range finder, infrared range-measurement system, reduce test
Measuring apparatus cost, improves real-time positioning measurement efficiency.
System of the present invention substitutes moving target with template and carries out auxiliary directional, range finding, using template calibrating camera
Derive and calculate initial parameter, detection template center position in the picture, i.e. template image space two-dimensional position, according to initial mark
Determining parameter and template image space two-dimensional coordinate system being mapped to three-dimensional coordinate system, Target space position is obtained with this (includes
Deflection and distance).Being determined to of Target space position overcome overlap and partial occlusion in the case of, moving target is in figure
The shortcoming that image plane can not be accurately identified, particularly in pedestrian detection, acoustic source array orientation during practical application, as beneficial
Auxiliary initial alignment parameter acquiring means, more highlight its engineering practical value.
Description of the drawings
Fig. 1 I types and II type calibrating template schematic diagrams;
Fig. 2 any directions equidistantly take a schematic diagram;
Fig. 3 central points completeness schematic diagram (n=1);
Fig. 4 central point O coding schematic diagrams;
Fig. 5 summits E coding schematic diagrams;
Fig. 6 camera calibration method schematic diagrams;
Fig. 7 Template Location schematic diagrams;
Fig. 8 camera calibration processes;
Fig. 9 target range 3m testing result schematic diagrams;
2 assignment test schematic diagram of Figure 10 examples;
3 assignment test schematic diagram of Figure 11 examples;
4 assignment test schematic diagram of Figure 12 examples;
5 assignment test schematic diagram of Figure 13 examples;
Figure 14 monocular vision target designation procedure declaration figures.
Specific embodiment
The present invention is explained in further detail with specific embodiment below in conjunction with the accompanying drawings:
A kind of monocular vision target designation method based on calibrating template,
The first step, defines calibrating template as requested, defines target step as follows:
(1) calibrating template is defined
As shown in figure 1, calibrating template is divided into I types in Fig. 1, two kinds of forms of II types, using diagonal black squares square
(two-square feature) is template primitive form, is defined as camera calibration formwork style.The I types or II pattern plates are only
Vertical use, black and white print on A4 paper (international standard size is 210mm × 297mm), it is desirable to horizontally and vertically placed in the middle, wherein,
If the length of side of any one black squares is l (units:Millimeter).
(2) completeness of central point and summit is defined
In Fig. 1, the O points of I types and II type calibrating templates are defined as the central point of calibrating template, afterwards abbreviation central point;A、B、
C, D, E and F point is defined as the summit of calibrating template, afterwards abbreviation summit.
According to template form, the four-quadrant subregion of central point is defined, as shown in Fig. 2 four subregions are designated as I respectively1、I2、
I3、I4, n pixel that any central point is equidistantly chosen on any i directions, its average isCorresponding central point is in I1
The pixel average of subregionFor:
The pixel average of central point four subregions is obtained in the same mannerCentral point completeness is defined as follows:
Complete central point:
Half complete central point:
Incomplete central point:Do not exist;
Wherein, TH1Represent the similarity degree of diagonal black bars or diagonal white portion pixel in image, TH2Represent figure
The difference degree of black and white area pixel as in.Summit completeness definition is with central point in the same manner.
In the present embodiment, n is to take a numberWhereinFor rounding downwards behaviour
Make.In Fig. 2, black takes a position with white expression, as a kind of embodiment, respectively with 45 °, 135 °, 225 ° and 315 ° directions
Take a little, count n=4.The pixel average of 4 subregions is designated as, Using said method, to determine each point
Completeness.
(3) summit is encoded with central point
In the present embodiment, using explanation as a example by I pattern plates.Pixel average on 4 subregions of calibrating template central point O point
Do not rememberAccording to the coding that pixel average determines 4 subregions, wherein black is with 0 coding, white with 1 volume
Code;
Determine each subregion pixel average in summitArbitrary summit is encoded, wherein black uses 0
Coding, white are encoded with 1;
Pixel average according to four subregions of central point and summit determines that its color is encoded.Central point and vertex encoding
Illustrate as shown in Figure 4 and Figure 5, encoder dictionary is as shown in table 1.
Second step, demarcates photographic head using template, determines initial alignment parameter betai
Calibrating template central point O points are positioned on camera lens group central axis, (or burnt flat parallel to lens plane
Face), position relationship is as shown in Figure 6.
Calibrating template is taken pictures sampling to the translation of remote continuous horizontal by near, obtain each apart from upper uncalibrated image, respectively to often
Uncalibrated image detects all angle points using Harris Corner Detection Algorithms, extracts complete central point, half complete central point, complete
Summit and half complete summit, according to the corresponding central point of vertex encoding coupling and summit, obtain central point and apex coordinate.With summit
O(x0,y0) and summit B (xB,yB) and summit E (xE,yE) as a example by, it is known that:
Wherein, dBOFor 2 points of Euclidean pixel distance of B, O in image, dEOFor 2 points of E, O in image Euclidean pixel away from
From dlEuclidean pixel distance between black patch summit and central point is represented, is then had:
dl=E (dBO, dEO)
Wherein, DKOEmpty with the Euclidean of camera lens central point K for image center (now, as template center's point O)
Between distance, thereby determine that initial alignment parameter betai:
3rd step:Calibrating template is fixed together with target to be measured, replaces target position to be measured using calibrating template position
Put, carry out image acquisition and video capture, as shown in Figure 7.Specifically, template is fixed on moving target center, by taking the photograph
As head gathers video image, and frame of video is converted to grayscale image sequence.
4th step:Using Harris Corner Detection Algorithms, angle point search, calculating are carried out to grey-level image frame.
5th step:The completeness of each angle point is detected, complete and half complete central point and summit is extracted.
6th step:Coupling of tabling look-up is carried out according to central point and vertex encoding table, determines central point O with summit B, C, E, F
Coordinate, taking one pair of which summit carries out Euclidean pixel distance calculating with central point, is designated as dl.
7th step:Determine image center P (x, y), the Euclidean pixel of calculation template central point O and image center P away from
From dOP, according to initial parameter βiWith dlDetermine image center with the Euclidean space of camera center apart from DKP.
8th step:According to dOPAnd DKP, calculation template center is with video camera apart from D and azimuth angle alpha.
Thus, the concrete grammar of step 4 to step 8 it is also possible to use following statement:
Using corner detection approach, whole angle points in input picture (collection image) or frame of video are detected, according to the complete of angle point
Standby property, inspection center's point and summit.According to coding schedule, coupling central point and each summit, central point and each apex coordinate, meter is obtained
Euclidean pixel distance d in nomogram picture between template central point and summitl.Input picture central point is designated as P (x, y), template center
The Euclidean pixel distance of point O and image center P is dOP, according to initial alignment parameter betaiEuclidean picture between central point and summit
Element is apart from dl, determine that image center P is D with the Euclidean space distance of camera centerKP.
DKP=βi*dl
Locating template center, i.e., be oriented to target, find range, according to DKPWith dOPDetermine template center's point and video camera
Euclidean space apart from D and azimuth angle alpha
The present embodiment solves to initialize location algorithm complexity in existing Technology for Target Location, it is often necessary to laser range finder,
The auxiliary positioning means such as infrared range-measurement system, testing cost are high, the problem of poor real.A kind of monocular vision based on template is provided
Object localization method.By monocular cam/video camera, moving target is replaced to carry out auxiliary directional, range finding, algorithm using template
Complexity is low, reduce test equipment cost, greatly strengthen target positioning real-time and accuracy, be succeeding target tracking,
The practical applications such as target detection, acoustic source array initial alignment provide necessary initiation parameter and real time calibration guarantee.
By adopting above-mentioned technical proposal, a kind of monocular vision target positioning side based on template that the present embodiment is provided
Method, has such beneficial effect compared with prior art:
Existing object localization method needs extra aiding sensors or distance-measuring equipment, needs to carry out coordinate system transformation or again
Modeling, computation complexity are high, it is impossible to provide accurate target component information for subsequent treatment in real time, cause target following, target
Identification, voice positioning etc. are processed and calculate failure, meanwhile, testing cost is too high, and multi-sensor information fusion also can reduce accuracy,
Higher to optimized algorithm requirement, it is unfavorable for that requirement of real-time height and the mini-plant of low-power consumption are adopted.
The present embodiment replaces moving target using template, template is detected, oriented and is found range, and is demarcated using template
Video camera, calculates initial alignment parameter, determines template center using Corner Detection and the detection of angle point completeness, according to initial ginseng
Two-dimensional image space template position is mapped to three-dimensional space position by number, obtains distance and direction of the moving target away from photographic head
Angle.The method can be efficiently applied to indoor and outdoor scene, it is adaptable to single or multiple moving target positioning, solve motion mesh
Mark easily by illumination, the such environmental effects such as block and cause the problem that feature extraction is difficult, detection is realized, computation complexity is low, energy
Enough fast and effectively positioning moving targets, the determination of moving target locus can be effectively applied to target following, pedestrian
In the subsequent treatment such as detection, acoustic source array orientation, with certain engineering use value.
As another kind of embodiment, corresponding to above-described embodiment in method, the present embodiment proposes a kind of to be based on calibration mold
The monocular vision target designation system of plate, including:
Calibrating template, for calibrating camera, is used for detecting calibrating template image space two-dimensional position;
Calibrating template image space two-dimensional coordinate system is mapped to three-dimensional coordinate system by coordinate mapping module, and
Derivation module, the projection according to calibrating template in the plane of delineation derive which in three dimensions with respect to photographic head
Range-azimuth is with to monocular vision target designation.
Further, the calibrating template, adopts diagonal black squares square for template primitive form, with central point
With summit, the completeness of central point and summit is defined, opposite vertexes are encoded with central point.
Further,
The derivation module, including:
Parameter determination module:Calibrating template calibrating camera is made, initial alignment parameter is determined;
Image capture module:Calibrating template and target bind to be measured, carry out image acquisition;
Computing module:
The collection image angle point is searched for, complete and half complete central point and summit is extracted, calculating gathers its in image
In a pair of central points and summit Euclidean pixel distance;
Determine the collection image center that monocular vision is collected;
According to the central point and summit Euclidean pixel distance of the calibrating template of the collection image, and initial alignment parameter,
Calculate the Euclidean pixel distance of collection image center and camera center;
Calculate the Euclidean pixel distance of collection image center and calibrating template central point;
According to above-mentioned collection image center and camera center Euclidean pixel distance and
Collection image center and the Euclidean pixel distance of calibrating template central point,
Calculate distance, the azimuth at calibrating template center and camera center.
Further, the parameter determination module:
Calibrating template central point O points are positioned on camera lens group central axis, parallel to lens plane;
Calibrating template is taken pictures sampling to the translation of remote continuous horizontal by near, obtain each apart from upper uncalibrated image, respectively to often
Open uncalibrated image and detect all angle points, extract complete central point, half complete central point, complete summit and half complete summit, according to
The corresponding central point of vertex encoding coupling and summit, obtain central point and apex coordinate;
Calculate the Euclidean pixel distance d of 2 points of B, O in uncalibrated imageBO, the Europe of 2 points of E, O in uncalibrated image
Family name pixel distance dEO:
Wherein, summit O (x0,y0), summit B (xB,yB), summit E (xE,yE);
dlRepresent Euclidean pixel distance between black patch summit and central point:
dl=E (dBO, dEO) (3)
Determine initial alignment parameter betai:
Wherein, DKOEuclidean space distance for uncalibrated image central point O and camera center point K.
Further, the computing module:
The central point of collection image is designated as P (x, y), according to central point and the summit Europe of the calibrating template of the collection image
Family name pixel distance dl, and initial alignment parameter betai, calculate the Euclidean space distance of collection image center P and camera center point K
DKP:
DKP=βi×dl(5)
Calibrating template central point O (x0,y0) with collection image center P (x, y) Euclidean pixel distance be dOP:
Wherein:(x0,y0) it is template central point O point coordinates in collection image.
The Euclidean space of calibrating template central point and camera center is apart from D and azimuth angle alpha:
Further, the central point of the calibrating template with the definition step of the completeness on summit is:
According to template form, the four-quadrant subregion on central point or summit is defined, four subregions are designated as I respectively1、I2、I3、I4,
The n pixel equidistantly chosen on any i directions by any central point or summit, its average isCorresponding central point or top
Point is in I1The pixel average of subregionFor:
In the same manner, the pixel average of four subregions of central point or summit is obtained with thisCentral point or top
Point completeness is defined as follows:
Complete central point or summit:
Half complete central point or summit:
Incomplete central point or summit:Do not exist;
Wherein, TH1Represent the similarity degree of diagonal black bars or diagonal white portion pixel in image, TH2Represent figure
The difference degree of black and white area pixel as in.
Further, the summit to calibrating template and comprising the concrete steps that central point is encoded:
Pixel average on 4 subregions of calibrating template central point O is remembered respectivelyEqual according to pixel
Value determines the coding of 4 subregions, and wherein black is with 0 coding, white with 1 coding;
Determine each subregion pixel average in summitArbitrary summit is encoded, wherein black uses 0
Coding, white are encoded with 1;
Pixel average according to four subregions of central point and summit determines that its color is encoded.
Embodiment 1:Camera calibration, initial alignment parameter determination
The present embodiment is camera calibration process, determines initial alignment parameter, and institute's extracting method of the present invention is tested.In room
In external environment, video camera is mounted and fixed at the top of certain robot or spider, and level shoots, and this is taken the photograph using cross grid
Camera, obtains automatically photographic head and shoots, and in the visual field of camera lens, the hand-held template of moving target is continuously put down between 0.5 meter -5 meters
Move, video camera from 0.5 meter of target range when start to take pictures, target range often increases by 0.5 meter of video camera and shoots once, shooting process
In, it is ensured that try one's best just to photographic head in template direction, it is ensured that template center's point keeps overlapping with video camera shooting image center.
Instance parameter explanation:Picture format PNG, picture size 1920 × 1080, uncalibrated image number 10.
This example calibration process as shown in figure 8, as a example by testing result when target range video camera 3m, as shown in figure 9,
Initial alignment parameter result of calculation is as shown in table 2, using initial parameter βiCalculate now measurement distance and deflection, with reality away from
From being contrasted with actual angle, obtain range error and angular error, due in calibration process template center all the time with shooting
Head center superposition, actual angle are 0 °.Calibrated error is as shown in table 3, it is known that, range error scope in 15mm, angular error
Scope is reached in error allowed band index in 1.2 °.
Embodiment 2:Monocular vision object localization method performance test based on template
The present embodiment, is carried out to target using monocular-camera based on template according to initial alignment parameter based on embodiment 1
Positioning.Indoors in environment, video camera is mounted and fixed at the top of certain robot or spider, and level shoots, this employing
Cross grid video camera, automatically obtains photographic head and shoots, in the visual field of camera lens, the hand-held template of moving target, according to set away from
From moving with direction, chi is demarcated using ground and determine target actual range and actual angle, using context of methods to shooting image
Detected, measured value and actual value technology distance error and angular error.
Instance parameter explanation:Picture format PNG, picture size 1920 × 1080, uncalibrated image number 8.This example is fixed
Position process is as shown in Figure 10, using initial parameter βiNow measurement distance and deflection is calculated, with actual range and actual angle
Contrasted, obtained range error and angular error, positioning result and method the performance test results are as shown in table 4, it is known that, distance
In 15mm, angular error scope reaches in error allowed band index in 1.2 ° range of error.
Embodiment 3:Corridor environment, single target motion conditions
The present embodiment applies the present invention under the static shooting state of video camera, single target motion positions in corridor.?
Under the conditions of this, video camera is mounted and fixed at the top of certain robot or spider, and level shoots, in the visual field of camera lens, one
The hand-held template of individual human target, template direction are tried one's best just to video camera, speed of the target according to 0.6m/s, within sweep of the eye by
Remote and near near video camera.The camera calibration process of this example Case-based Reasoning 1, positions to moving target in video.
Embodiment parameter declaration:Video format MP4,160 frame of video frame number, video image size 1920 × 1080.
This example position fixing process as shown in figure 11, with the 10th frame, 30 frames, 50 frames, 70 frames, 90 frames, 110 frames, 130 frames, 150
As a example by frame, moving target motor process and testing result are described, target positioning result is as shown in table 5.
Embodiment 4:Outdoor environment, single target motion conditions
The present embodiment applies the present invention under the static shooting state of video camera, and in outdoor square, single target motion is fixed
Position.With this understanding, video camera is mounted and fixed at the top of certain robot or spider, and level shoots, in the visual field of camera lens
Interior, the hand-held template of human target, template direction are tried one's best just to video camera, and speed of the target according to 0.6m/s, in visual field model
Draw near away from video camera in enclosing.The camera calibration process of this example Case-based Reasoning 1, it is fixed that moving target in video is carried out
Position.
Embodiment parameter declaration:Video format MP4,160 frame of video frame number, video image size 1920 × 1080.
This example position fixing process as shown in figure 12, with the 10th frame, 30 frames, 50 frames, 70 frames, 90 frames, 110 frames, 130 frames, 150
As a example by frame, moving target motor process and testing result are described.Target positioning result is as shown in table 6.
Embodiment 5:Outdoor environment, two target motion conditions
The present embodiment applies the present invention under the static shooting state of video camera, and in outdoor square, two moving targets are fixed
Position.With this understanding, video camera is mounted and fixed at the top of certain robot or spider, and level shoots, in the visual field of camera lens
Interior, 1 trumpeter of target holds I pattern plates, and 2 trumpeter of target holds II pattern plates, and template direction is tried one's best just to video camera, respectively according to
The speed of 0.6m/s, in the video camera that draws near within sweep of the eye.The camera calibration process of this example Case-based Reasoning 1,
Moving target in video is positioned.
Embodiment parameter declaration:Video format MP4,160 frame of video frame number, video image size 1920 × 1080.
This example position fixing process as shown in figure 12, with the 10th frame, 30 frames, 50 frames, 70 frames, 90 frames, 110 frames, 130 frames, 150
As a example by frame, moving target motor process and testing result are described, target positioning result is as shown in table 5.
Subordinate list:
1 I patterns plate central point of table-vertex encoding table;
2 initial alignment parameter list of table;
3 initial alignment assignment test table of table;
4 example of table, 2 positioning result;
5 example of table, 3 positioning result;
6 example of table, 4 positioning result;
7 example of table, 5 positioning result.
1 I patterns plate central point of table-vertex encoding table
2 initial alignment parameter list of table
3 initial alignment of table positions table
4 example of table, 2 positioning result
5 example of table, 3 positioning result
6 example of table, 4 positioning result
7 example of table, 5 positioning result
The above, only the invention preferably specific embodiment, but the protection domain of the invention is not
Be confined to this, any those familiar with the art in the technical scope that the invention is disclosed, according to the present invention
The technical scheme of creation and its inventive concept equivalent or change in addition, should all cover the invention protection domain it
Interior.
Claims (7)
1. a kind of monocular vision target designation system based on calibrating template, it is characterised in that include:
Calibrating template, for calibrating camera, is used for detecting calibrating template image space two-dimensional position;
Calibrating template image space two-dimensional coordinate system is mapped to three-dimensional coordinate system by coordinate mapping module, and
Derivation module, the projection according to calibrating template in the plane of delineation derive which in three dimensions with respect to the distance of photographic head
With orientation with to monocular vision target designation.
2. the monocular vision target designation system based on calibrating template as claimed in claim 1, it is characterised in that the demarcation
Template, adopts diagonal black squares square for template primitive form, with central point and summit, defines central point with summit
Completeness, opposite vertexes are encoded with central point.
3. the monocular vision target designation system based on calibrating template as claimed in claim 1, it is characterised in that the derivation
Module, including:
Parameter determination module:Calibrating template calibrating camera is made, initial alignment parameter is determined;
Image capture module:Calibrating template and target bind to be measured, carry out image acquisition;
Computing module:
The collection image angle point is searched for, complete and half complete central point and summit is extracted, wherein in collection image is calculated
To central point and the Euclidean pixel distance on summit;
Determine the collection image center that monocular vision is collected;
According to the central point and summit Euclidean pixel distance of the calibrating template of the collection image, and initial alignment parameter,
Calculate the Euclidean pixel distance of collection image center and camera center;
Calculate the Euclidean pixel distance of collection image center and calibrating template central point;
According to above-mentioned collection image center and camera center Euclidean pixel distance and
Collection image center and the Euclidean pixel distance of calibrating template central point,
Calculate distance, the azimuth at calibrating template center and camera center.
4. the monocular vision target designation system based on calibrating template as claimed in claim 3, it is characterised in that the parameter
Determining module:
Calibrating template central point O points are positioned on camera lens group central axis, parallel to lens plane;
Calibrating template is taken pictures sampling to the translation of remote continuous horizontal by near, obtain each apart from upper uncalibrated image, respectively to per marking
Determine all angle points of image detection, extract complete central point, half complete central point, complete summit and half complete summit, according to summit
The corresponding central point of codes match and summit, obtain central point and apex coordinate;
Calculate the Euclidean pixel distance d of 2 points of B, O in uncalibrated imageBO, the Euclidean pixel distance d of 2 points of E, O in uncalibrated imageEO:
Wherein, summit O (x0,y0), summit B (xB,yB), summit E (xE,yE);
dlRepresent Euclidean pixel distance between black patch summit and central point:
dl=E (dBO, dEO) (3)
Determine initial alignment parameter betai:
Wherein, DKOEuclidean space distance for uncalibrated image central point O and camera center point K.
5. the monocular vision target designation system based on calibrating template as claimed in claim 3, it is characterised in that the calculating
Module:
The central point of collection image is designated as P (x, y), according to the central point and summit Euclidean picture of the calibrating template of the collection image
Element is apart from dl, and initial alignment parameter betai, the Euclidean space of collection image center P and camera center point K is calculated apart from DKP:
DKP=βi×dl(5)
Calibrating template central point O (x0,y0) with collection image center P (x, y) Euclidean pixel distance be dOP:
Wherein:(x0,y0) it is template central point O point coordinates in collection image.
The Euclidean space of calibrating template central point and camera center is apart from D and azimuth angle alpha:
6. the monocular vision target designation system based on calibrating template as claimed in claim 2, it is characterised in that the demarcation
The central point of template with the definition step of the completeness on summit is:
According to template form, the four-quadrant subregion on central point or summit is defined, four subregions are designated as I respectively1、I2、I3、I4, to appointing
The n pixel that meaning central point or summit are equidistantly chosen on any i directions, its average isCorresponding central point or summit exist
I1The pixel average of subregionFor:
In the same manner, the pixel average of four subregions of central point or summit is obtained with thisCentral point or summit are complete
Standby property is defined as follows:
Complete central point or summit:
Half complete central point or summit:
Incomplete central point or summit:Do not exist;
Wherein, TH1Represent the similarity degree of diagonal black bars or diagonal white portion pixel in image, TH2Represent in image
The difference degree of black and white area pixel.
7. the monocular vision target designation system based on calibrating template as claimed in claim 2, it is characterised in that described pair of mark
The summit of solid plate and comprising the concrete steps that central point is encoded:
Pixel average on 4 subregions of calibrating template central point O is remembered respectivelyDetermined according to pixel average
The coding of 4 subregions, wherein black are encoded with 1 with 0 coding, white;
Determine each subregion pixel average in summitArbitrary summit is encoded, wherein black is encoded with 0,
White is encoded with 1;
Pixel average according to four subregions of central point and summit determines that its color is encoded.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610910942.2A CN106504287B (en) | 2016-10-19 | 2016-10-19 | Monocular vision object space positioning system based on template |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610910942.2A CN106504287B (en) | 2016-10-19 | 2016-10-19 | Monocular vision object space positioning system based on template |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106504287A true CN106504287A (en) | 2017-03-15 |
CN106504287B CN106504287B (en) | 2019-02-15 |
Family
ID=58294328
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610910942.2A Expired - Fee Related CN106504287B (en) | 2016-10-19 | 2016-10-19 | Monocular vision object space positioning system based on template |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106504287B (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108761436A (en) * | 2018-08-27 | 2018-11-06 | 上海岗消网络科技有限公司 | A kind of flame visual token device and method |
CN108828965A (en) * | 2018-05-24 | 2018-11-16 | 联想(北京)有限公司 | Localization method, electronic equipment and smart home system, storage medium |
CN109636859A (en) * | 2018-12-24 | 2019-04-16 | 武汉大音科技有限责任公司 | A kind of scaling method of the 3D vision detection based on one camera |
CN109887025A (en) * | 2019-01-31 | 2019-06-14 | 沈阳理工大学 | Monocular self-adjustable fire point 3-D positioning method and device |
CN113538578A (en) * | 2021-06-22 | 2021-10-22 | 恒睿(重庆)人工智能技术研究院有限公司 | Target positioning method and device, computer equipment and storage medium |
CN114018215A (en) * | 2022-01-04 | 2022-02-08 | 智道网联科技(北京)有限公司 | Monocular distance measuring method, device, equipment and storage medium based on semantic segmentation |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103487034A (en) * | 2013-09-26 | 2014-01-01 | 北京航空航天大学 | Method for measuring distance and height by vehicle-mounted monocular camera based on vertical type target |
CN103578133A (en) * | 2012-08-03 | 2014-02-12 | 浙江大华技术股份有限公司 | Method and device for reconstructing two-dimensional image information in three-dimensional mode |
CN104484883A (en) * | 2014-12-24 | 2015-04-01 | 河海大学常州校区 | Video-based three-dimensional virtual ship positioning and track simulation method |
TWM505820U (en) * | 2015-04-22 | 2015-08-01 | Waitemata Hands Co Ltd | Improved structure of socks |
CN106651957A (en) * | 2016-10-19 | 2017-05-10 | 大连民族大学 | Monocular vision target space positioning method based on template |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2003530039A (en) * | 2000-04-04 | 2003-10-07 | コーニンクレッカ フィリップス エレクトロニクス エヌ ヴィ | Automatic calibration of panorama / tilt / zoom cameras |
-
2016
- 2016-10-19 CN CN201610910942.2A patent/CN106504287B/en not_active Expired - Fee Related
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103578133A (en) * | 2012-08-03 | 2014-02-12 | 浙江大华技术股份有限公司 | Method and device for reconstructing two-dimensional image information in three-dimensional mode |
CN103487034A (en) * | 2013-09-26 | 2014-01-01 | 北京航空航天大学 | Method for measuring distance and height by vehicle-mounted monocular camera based on vertical type target |
CN104484883A (en) * | 2014-12-24 | 2015-04-01 | 河海大学常州校区 | Video-based three-dimensional virtual ship positioning and track simulation method |
TWM505820U (en) * | 2015-04-22 | 2015-08-01 | Waitemata Hands Co Ltd | Improved structure of socks |
CN106651957A (en) * | 2016-10-19 | 2017-05-10 | 大连民族大学 | Monocular vision target space positioning method based on template |
Non-Patent Citations (1)
Title |
---|
周娜: "基于单目视觉的摄像机定位技术研究", 《中国优秀硕士学位论文全文数据库》 * |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108828965A (en) * | 2018-05-24 | 2018-11-16 | 联想(北京)有限公司 | Localization method, electronic equipment and smart home system, storage medium |
CN108761436A (en) * | 2018-08-27 | 2018-11-06 | 上海岗消网络科技有限公司 | A kind of flame visual token device and method |
CN108761436B (en) * | 2018-08-27 | 2023-07-25 | 上海岗消网络科技有限公司 | Flame vision distance measuring device and method |
CN109636859A (en) * | 2018-12-24 | 2019-04-16 | 武汉大音科技有限责任公司 | A kind of scaling method of the 3D vision detection based on one camera |
CN109636859B (en) * | 2018-12-24 | 2022-05-10 | 武汉大音科技有限责任公司 | Single-camera-based calibration method for three-dimensional visual inspection |
CN109887025A (en) * | 2019-01-31 | 2019-06-14 | 沈阳理工大学 | Monocular self-adjustable fire point 3-D positioning method and device |
CN113538578A (en) * | 2021-06-22 | 2021-10-22 | 恒睿(重庆)人工智能技术研究院有限公司 | Target positioning method and device, computer equipment and storage medium |
CN114018215A (en) * | 2022-01-04 | 2022-02-08 | 智道网联科技(北京)有限公司 | Monocular distance measuring method, device, equipment and storage medium based on semantic segmentation |
CN114018215B (en) * | 2022-01-04 | 2022-04-12 | 智道网联科技(北京)有限公司 | Monocular distance measuring method, device, equipment and storage medium based on semantic segmentation |
Also Published As
Publication number | Publication date |
---|---|
CN106504287B (en) | 2019-02-15 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106504287A (en) | Monocular vision object space alignment system based on template | |
CN106651957B (en) | Monocular vision object space localization method based on template | |
CN104173054B (en) | Measuring method and measuring device for height of human body based on binocular vision technique | |
CN104200086B (en) | Wide-baseline visible light camera pose estimation method | |
CN102788559B (en) | Optical vision measuring system with wide-field structure and measuring method thereof | |
CN111340797A (en) | Laser radar and binocular camera data fusion detection method and system | |
CN103424112B (en) | A kind of motion carrier vision navigation method auxiliary based on laser plane | |
Tamas et al. | Targetless calibration of a lidar-perspective camera pair | |
CN102609941A (en) | Three-dimensional registering method based on ToF (Time-of-Flight) depth camera | |
CN104142157A (en) | Calibration method, device and equipment | |
CN103759669A (en) | Monocular vision measuring method for large parts | |
CN113592989A (en) | Three-dimensional scene reconstruction system, method, equipment and storage medium | |
CN104021588A (en) | System and method for recovering three-dimensional true vehicle model in real time | |
CN102136140B (en) | Rectangular pattern-based video image distance detecting method | |
CN104034269A (en) | Monocular vision measuring method and monocular vision measuring device | |
CN112884841B (en) | Binocular vision positioning method based on semantic target | |
CN102930551B (en) | Camera intrinsic parameters determined by utilizing projected coordinate and epipolar line of centres of circles | |
CN103852060A (en) | Visible light image distance measuring method based on monocular vision | |
CN108469254A (en) | A kind of more visual measuring system overall calibration methods of big visual field being suitable for looking up and overlooking pose | |
CN110415299B (en) | Vehicle position estimation method based on set guideboard under motion constraint | |
CN111998862A (en) | Dense binocular SLAM method based on BNN | |
Yuan et al. | Combining maps and street level images for building height and facade estimation | |
CN103985121B (en) | Method for optical calibration of underwater projector structure | |
CN114066985B (en) | Method for calculating hidden danger distance of power transmission line and terminal | |
CN109035343A (en) | A kind of floor relative displacement measurement method based on monitoring camera |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20190215 Termination date: 20211019 |
|
CF01 | Termination of patent right due to non-payment of annual fee |