CN110480180A - The object recognition and detection method of foreign matter device is removed based on reflection-type laser - Google Patents
The object recognition and detection method of foreign matter device is removed based on reflection-type laser Download PDFInfo
- Publication number
- CN110480180A CN110480180A CN201810624639.5A CN201810624639A CN110480180A CN 110480180 A CN110480180 A CN 110480180A CN 201810624639 A CN201810624639 A CN 201810624639A CN 110480180 A CN110480180 A CN 110480180A
- Authority
- CN
- China
- Prior art keywords
- foreign matter
- laser
- coordinate
- image
- point
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B23—MACHINE TOOLS; METAL-WORKING NOT OTHERWISE PROVIDED FOR
- B23K—SOLDERING OR UNSOLDERING; WELDING; CLADDING OR PLATING BY SOLDERING OR WELDING; CUTTING BY APPLYING HEAT LOCALLY, e.g. FLAME CUTTING; WORKING BY LASER BEAM
- B23K26/00—Working by laser beam, e.g. welding, cutting or boring
- B23K26/02—Positioning or observing the workpiece, e.g. with respect to the point of impact; Aligning, aiming or focusing the laser beam
- B23K26/03—Observing, e.g. monitoring, the workpiece
- B23K26/032—Observing, e.g. monitoring, the workpiece using optical means
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B23—MACHINE TOOLS; METAL-WORKING NOT OTHERWISE PROVIDED FOR
- B23K—SOLDERING OR UNSOLDERING; WELDING; CLADDING OR PLATING BY SOLDERING OR WELDING; CUTTING BY APPLYING HEAT LOCALLY, e.g. FLAME CUTTING; WORKING BY LASER BEAM
- B23K26/00—Working by laser beam, e.g. welding, cutting or boring
- B23K26/36—Removing material
- B23K26/38—Removing material by boring or cutting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Optics & Photonics (AREA)
- Plasma & Fusion (AREA)
- Mechanical Engineering (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Length Measuring Devices By Optical Means (AREA)
Abstract
The invention discloses a kind of object recognition and detection methods that foreign matter device is removed based on reflection-type laser.For the feature of cable winds foreign matter, by customized image recognition algorithm, cable diameter width and foreign matter winding point coordinate in image are obtained;Actual distance is obtained according to corresponding relationship;The plane of delineation coordinate of laser point is calculated according to actual distance;The angle coordinate of laser point and foreign matter point is converted to angle coordinate system by plane of delineation coordinate system, is positioned.The discrimination in laser removing foreign matter device operation process to target is improved using method of the invention, can also be accurately tracked when target is shaken, so that laser is accurately cut foreign matter, improve laser cutting efficiency.
Description
Technical field
The invention belongs to power transmission cable foreign matters to remove tracking field, specifically, being a kind of based on the removing of reflection-type laser
The object recognition and detection method of foreign matter device.
Background technique
Research to video image mainly includes two aspects of detection and tracking of target.In general, one it is complete
Video frequency following system will pass through video image acquisition, target detection and independently track these steps, finally according to tracking result
Carry out the operations such as other analyses or control.
In recent years, gradually go deep into the related fieldss research such as computer vision, image procossing and characteristic matching etc. are calculated
The precision and real-time of method are continuously improved.
The target of visual servo detection is exactly that reliable and stable visual feedback information is provided for visual servo process, because
This, needs high-efficient, fireballing image procossing or parser.It has been directed to Image Acquisition or video captures, filtering is gone
Make an uproar, image segmentation, feature extraction and feature description etc. many aspects.The Feature extraction and recognition of plurality of target under complex environment
It is particularly important for the performance of vision servo system.
The method of Feature extraction and recognition is more, however there is no a kind of feature extractions suitable under any scene at present
Method, and also ununified recognition methods, most Feature extraction and recognitions are all based under certain task or scene
's.It needs to select readily identified in the feature of selection target and extract, the feature that strong operability and repetitive rate are high.Visual signature
Quantity system performance is had a major impact, target signature excessively will lead to calculation amount increase, influence the spirit of vision servo system
Sensitivity;Target signature is very few and to will lead to the calculating of vision servo system unreliable, cannot reach good precision.It should select simultaneously
Suitable visual signature is taken, to guarantee the nonsingularity of image turn.It handles huge characteristic set and is unable to satisfy figure
As the requirement of real-time of processing, therefore the quantity of feature need to be minimized under the premise of guaranteeing that characteristic set is sensitive reliable, and be somebody's turn to do
Freedom degree and used control method of the minimal characteristic set dependent on vision servo system control closed loop.Local feature is answered
With extensive, and the control effect of available satisfaction under normal conditions, however local image characteristics are mainly for known rule
The then target of shape causes its limitation, and works as visual signature and exceed camera coverage, can not obtain complete characteristic information
When, system would become hard to make accurate operation.Especially in true environment, illumination condition, target occlusion, ambient noise and object
Shape, texture etc. the numerical precision of visual signature can be impacted, and then influence the execution performance of visual servo task.
Global image feature can make up the deficiency of local geometric features applied to visual servo.Common global characteristics have image moment,
Fourier descriptor, image brightness properties etc..The most global image feature of application is image moment characteristics at present.However, global
The positioning accuracy of characteristics of image is lower than local geometric features, and sensitive to background noise.
As it can be seen that general method is not present in the selection of visual signature, the working environment according to robot is needed, in system reality
Comprehensively considered between Shi Xing, complexity and stability.
Summary of the invention
The present invention proposes a kind of object recognition and detection method that foreign matter device is removed based on reflection-type laser, twines for cable
Around the feature of foreign matter, the discrimination in laser removing foreign matter device operation process to target is effectively improved, when target is shaken
Also it can accurately track, laser is enable accurately to cut foreign matter, improve laser cutting efficiency.
The technical solution for realizing the aim of the invention is as follows: a kind of target knowledge for removing foreign matter device based on reflection-type laser
Other localization method, comprising the following steps:
Step 1: visual sensor acquires the current image frame including foreign matter target, and is transmitted to control system, aobvious
Real-time display image in display screen;
Step 2: after control system receives the incoming trace command of input unit, image recognition is carried out;If identification is lost
It loses, return step one;Otherwise, width d that cable diameter projects in visual sensor plane and different is obtained by image recognition
The image coordinate of object winding point;
Step 3: the actual distance Zc of foreign matter distance means is found out according to width d;
Step 4: coordinate of the laser point in the plane of delineation is calculated as coordinates of targets (x according to distance Zc0,y0);
Step 5: calculating coordinate of the plane of delineation coordinate system to angle coordinate system;Control system is according to vision under distance Zc
The corresponding relationship of pixel and angle coordinate system in sensor plane coordinate system obtains the angle coordinate (θ of laser pointx0,θy0) and
Angle coordinate (the θ of foreign matter pointx1,θy1), by foreign matter point location;
Step 6: it is poor to make to the angle coordinate of laser point and foreign matter point, obtains angle delta θ=(Δ θ that turntable should rotatex,
Δθy), control system sends commands to turntable, is overlapped laser point and foreign matter point.
Further, for step 2, specific steps are as follows:
Step 2.1: the pixel value that control system captures visual sensor is the image gray processing of n*m.
Step 2.2: image being converted to the form of byte array, an available n*m is (at this time for 1920*1080's)
Image array, wherein each elements AijThe gray value of the corresponding pixel.
Step 2.3: obtaining the image information for specifying region in image, be converted to image recognition matrix, matrix is indicated
Image information carries out the projection of x-axis and y-axis respectively in the coordinate system of visual sensor plane;For every in image recognition matrix
A line, each column seek ∑ Ai,j, curve graph is done to the result, obtains the projection in x-axis and y-axis.
Step 2.4: by analyzing the obtained curve of projection, the corresponding image coordinate of available valley value, i.e. foreign matter is twined
Around a coordinate (x1,y1)。
Step 2.5: finding out in y-axis that the pixel of trough is wide, it can obtain cable diameter and thrown in visual sensor plane
The width d of shadow, cable true diameter are d0Rice, and it is known.
Further, for step 3, concrete operations are as follows: similar with triangle according to the pin-hole model in camera model
Principle obtains formula:Wherein, f is visual sensor focal length, and d is that cable diameter projects in visual sensor plane
Width, d0It is cable true diameter, the actual distance Zc of foreign matter distance means is found out by the formula.
Further, for step 4, specific steps are as follows:
Step 4.1: absolute parallel according to optical path and visual sensor imaging axle center, it is known that laser point is in visual sensor axis
At the fixed position relative of heart corresponding points, known to this relative position.In real coordinate system, if visual sensor axle center corresponding points
Coordinate is (ac,bc), then laser spot position is (ac+dx,bc+dy).Dx and dy is respectively laser point and vision under same distance
Fixation alternate position spike of the sensor axle center corresponding points under real coordinate system.
Step 4.2: the corresponding point in axle center is imaged in imaging process in visual sensor, no matter with a distance from visual sensor
It is how many, the coordinate in visual sensor plane is constant always, is (xc,yc).If laser point is in visual sensor plane
Coordinate be (x0,y0)。(x0,y0)=(xc+dx’,yc+dy').Dx ' and dy ' is laser point and visual sensor axle center corresponding points
Coordinate difference under visual sensor plane coordinate system.Then:
Solve (x0,y0)。
Further, for step 5, specific steps are as follows:
Calculating coordinate of the plane of delineation coordinate system to angle coordinate system.Control system descends visual sensor flat according to this distance
Pixel and the corresponding relationship of angle coordinate system obtain the angle coordinate of laser point and foreign matter point in areal coordinate system, and foreign matter point is determined
Position.
Step 5.1: resolving laser point relative angle coordinate (θx0,θy0)。
Step 5.2: resolving foreign matter point relative angle coordinate (θx1,θy1)。
Removing foreign matter device using the reflection-type laser of the above method includes turntable, laser part, input/output unit, control
System and visual sensor processed;Wherein, turntable includes orientation component and pitch components, and laser part includes laser beam emitting head, anti-
Penetrate mirror element one, reflection mirror component two;
Orientation component is mounted on above fixed frame, and pitch components are mounted on above orientation component, and laser beam emitting head is mounted on
In orientation component, reflection mirror component one, reflection mirror component two are mounted in pitch components, and visual sensor is mounted on pitch components
In;
Control system is connect with visual sensor, adjusts visual sensor focal length and wide-angle parameter;Visual sensor is used for
Foreign matter point is captured, image is transmitted to control system and carries out image recognition;Control system respectively with orientation component, pitch components phase
Even, control system handles the result of image recognition, and driving orientation component, pitch components carry out orientation and pitching two freely
The rotation of degree, so that visual sensor and laser part be driven to rotate;Input/output unit is connect with control system, for showing
The control interface and input operation information of control system.
Compared with prior art, the present invention its remarkable advantage is: (1) being directed to the feature of cable winds foreign matter, effectively
Improve the discrimination in laser removing foreign matter device operation process to target.(2) it can also be accurately tracked when target is shaken, make to swash
Light can accurately cut foreign matter, improve laser cutting efficiency.
Detailed description of the invention
Fig. 1 is the general assembly structural schematic diagram that reflection-type laser of the invention removes foreign matter device.
Fig. 2 is that reflection-type laser of the invention removes the orientation component of foreign matter device and the side view of pitch components.
Fig. 3 is the sectional view and side view for the turntable that reflection-type laser of the invention removes foreign matter device.
Fig. 4 is the top view and sectional view for the orientation component that reflection-type laser of the invention removes foreign matter device.
Fig. 5 is the front view and side view for the pitch components that reflection-type laser of the invention removes foreign matter device.
Fig. 6 is the control system operation schematic diagram that reflection-type laser of the invention removes foreign matter device.
Fig. 7 is target identification and localization method flow chart of the invention.
Fig. 8 is the result figure of collected complete image gray processing.
Fig. 9 is plane of delineation coordinate system schematic diagram.
Figure 10 is plane of delineation coordinate system about Y-axis projection ∑ Ai,jCurve.
Figure 11 is plane of delineation coordinate system about X-axis projection ∑ Ai,jCurve.
Figure 12 is pinhole camera modeling.
Figure 13 is the pinhole camera modeling after translation.
Specific embodiment
The invention will be further described with reference to the accompanying drawings and examples.
As shown in Figure 1, the general assembly structural schematic diagram of foreign matter device is removed for reflection-type laser used in the method for the present invention,
The device includes turntable 22, dust proof member 3, laser part 4, fixed frame 5 (preferably tripod), power-supply system, input and output dress
It sets, control system, communication module, visual sensor 10 (preferably high definition industrial camera);Wherein, turntable 22 include orientation component 1,
Pitch components 2.Orientation component 1 is mounted on 5 top of tripod;Pitch components 2 are mounted on 1 top of orientation component;In dust proof member 3
17 sets of dust cover outside pitch components 2, dust-proof mirror 31 is mounted in orientation component 1;Laser part 4 includes laser beam emitting head
19, reflection mirror component 1, reflection mirror component 2 21, laser beam emitting head 19 are mounted in orientation component 1, reflection mirror component one
20, reflection mirror component 2 21 is mounted in pitch components 2, and by this strapdown structure, optical path can pass through reflection mirror component
1 and reflection mirror component 2 21 successive reflex after, be parallel to the sending of 10 optical axis of visual sensor, therefore only need to adjust anti-
The angle for penetrating mirror, without making laser beam emitting head 19 that turntable 22 be followed to rotate;Visual sensor 10 is mounted in pitch components 2.
In control system, industrial personal computer 35 is connect with visual sensor 10, adjusts 10 focal length of visual sensor and wide-angle parameter,
Make to get a clear view;Visual sensor 10 and industrial personal computer 35 connect, and after capturing foreign matter point, image are transmitted to industrial personal computer 35 and carries out image
Identification;Industrial personal computer 35 is connected by controlling driving part 38 with orientation component 1, pitch components 2, can carry out orientation and pitching two
The rotation of freedom degree, so that laser part 4 and visual sensor 10 be driven to rotate;Industrial personal computer 35 is by the result treatment of image recognition
Afterwards, it is transmitted to control driving part 38, orientation component 1 is adjusted and pitch components 2 is tracked, and drive laser part 4 to foreign matter
It is cut;Control system and the connection of communication module 9 are for obtaining remote control information;Power-supply system is control system, vision
Sensor 10, orientation component 1, pitch components 2, laser part 4, communication module 9 are powered.
As shown in Figure 2-5, orientation component 1 includes orientation torque motor 11, (rotation becomes transmitter to orientation angles measuring device
Or Lineside encoding unit) 12, orientation limiting lock components 23, laser installation joint unit 24, pedestal 25 and orientation bracket 26.Its
In, pedestal 25 is fixed on 5 top of tripod;Orientation bracket 26 is mounted on pedestal 25 and pitch components 2 connect.Laser installation
Joint unit 24 for connecting laser part 4, and passes through laser in the circular hole in 25 center of pedestal;Orientation angles measurement
Device (rotation becomes transmitter or Lineside encoding unit) 12 is placed in inside pedestal 25, is obtained the angle of true bearing rotation by it, is led to
It crosses electric signal and is sent to control driving part 38;Orientation torque motor 11 be located at orientation angles measuring device (rotation become transmitter or
Person's Lineside encoding unit) 12 top, it can be achieved that orientation component 1 in orientation -180~180 degree rotation;Orientation limiting lock components
23 are located on 26 edge of orientation bracket, and the order by controlling driving part 38 makes it fall into a circular hole at 25 edge of pedestal
In, for locking orientation, constant bearing bracket 26.
Pitch components 2 include bearing block 1, bearing block 2 14, pitching moment motor 15, pitch axis 1, connection parts
28, pitch axis 2 29, pitch angle measuring device (rotation becomes transmitter or Lineside encoding unit) 16, pitching limiting lock components
30.Wherein, bearing block 1 and bearing block 2 14 are fixedly mounted on the orientation bracket 26 of orientation component 1, one 13 He of bearing block
Bearing block 2 14 is parallel to each other, and orientation component 1 is followed to rotate;Pitching moment motor 15 is located in bearing block 1, it can be achieved that 0~
90 degree of pitching rotations.Pitch axis 1 is located in bearing block 1, pitching moment motor 15 is connected, as pitching moment motor 15
Drive mechanism;28 crossbearer of connection parts is connected, as biography among pitch axis 1 and pitch axis 2 29 with pitch axis 1
Dynamic structure, and be connected with visual sensor 10, it rotates together;Pitch axis 2 29 is located in bearing block 2 14, connects connection parts
28, as driven structure;Pitch angle measuring device (rotation becomes transmitter or Lineside encoding unit) 16 is placed in bearing block 2 14,
Pitch angle measuring device (rotation becomes transmitter or Lineside encoding unit) 16 is able to detect the angle of the rotation of pitch axis 2 29, passes through
Electric signal is sent to control driving part 38, is further transmitted to control system 8;Pitching limiting lock components 30 are located at pitch axis
2 29 outside, the pitch angle for mechanical retention pitching moment motor 15.Orientation bracket 26 is fixed on pedestal 25, is used for
Pitch components 2 are connected, it is driven to rotate in orientation.
Dust proof member 3 includes dust cover 17, laser work window 18, dust-proof mirror 31.Dust cover 17 is covered in pitch components 2
Outside, for internal dust-proof;Laser work window 18 is used to pass through dust cover 17 when laser work in dust cover 17;It is dust-proof
Mirror 31 is located at 1 lower section of orientation component, at the circular hole that laser installation joint unit 24 connects, prevents impurity from entering.
Laser part 4 includes laser beam emitting head 19, reflection mirror component 1, reflection mirror component 2 21, laser controlling case 32.
Wherein, laser beam emitting head 19 is connect by optical fiber with laser controlling case 32, is emitted laser, is at laser light source;Laser beam emitting head
19 are connected with laser installation joint unit 24, together in the circular hole in 25 center of the pedestal of orientation component 1, make laser beam emitting head
19 straight up, laser by emitting head issue after straight up;Reflection mirror component 2 21 installs 24 phase of joint unit with laser
Even, it is located at right above laser beam emitting head 19, laser after emitting head sending on reflection mirror component 2 21 by straight up, passing through
45 ° of reflected illuminations are on reflection mirror component 1;Reflection mirror component 1 is mounted on the interior of the pitch axis 2 29 of pitch components 2
Side, after laser reflects away on reflection mirror component 1, light beam is parallel with 10 optical axis of visual sensor;Also, mirror part
Part 1 is rotated together with pitch axis 2 29 and visual sensor 10, can reflect the laser light to 10 observation area of visual sensor
In required position, to alleviate the burden of motor improve motor without rotating the biggish laser beam emitting head of weight
The precision of rotation.
Power-supply system includes laser battery 33 and DC power supply 34.DC power supply 34 is mainly used for negative to entire motor
It carries, input/output unit and visual sensor 10 are powered.Laser battery 33 is mainly used for the laser controlling in laser part 4
Case 32 is powered.
Input/output unit includes display 36 and input unit 37.Display 36 receives industrial personal computer 35 by HDMI interface
Signal, 8 control interface of display control program;Input unit 37 is used to input the operation information of operator, is sent to industrial personal computer
35。
As shown in fig. 6, control system includes industrial personal computer 35 and control driving part 38.Industrial personal computer 35 passes through RS485 serial ports
Visual sensor 10 is sent commands to, is changed for realizing focal length, the narrow angle of wide-angle, presetting bit image parameter;Industrial personal computer 35 passes through
Coaxial cable receives 10 image of visual sensor and information, identifies route and foreign matter, obtains foreign matter location information and feed back to control
Driving part 38;Industrial personal computer 35 is connect with control driving part 38 by RS422 serial ports and realizes that the rotation of turntable 22 is cut;Industrial personal computer
35 connect realization human-computer interaction with input/output unit;Control driving part 38 is located in the pedestal 25 in orientation component 1, passes through
Orientation angles measuring device (rotation become transmitter or Lineside encoding unit) 12, pitch angle measuring device (rotation become transmitter or
Lineside encoding unit) 16 angle informations for receiving orientation component 1 and pitch components 2 respectively, and industrial personal computer is transmitted to by RS422 serial ports
35, while 35 information of industrial personal computer, control azimuth component 1 and pitch components 2 are received, carry out two-freedom movement.
Communication module includes wireless transmitter module and wireless receiving module, is mainly used for realizing 35 wireless monitor function of industrial personal computer
Energy;Communication module has the function of wireless awakening and networking function.
Visual sensor 10 is fixedly mounted on 28 lower section of connection parts of pitch components 2, with rotating therewith, and in real time
It acquires image information in front and control system 8 is sent to by coaxial cable, and receive 35 signal instruction of industrial personal computer.
Tripod 5 is located at the bottommost of device, by changing the length adjustment device height of three feet, provides and stablizes properly
Shooting state.
In conjunction with Fig. 7, the present invention is based on the object recognition and detection methods that laser removes foreign matter device, comprising the following steps:
Step 1: visual sensor acquires the current image frame including foreign matter target, and is transmitted to control system, aobvious
Real-time display image in display screen;
Step 2: after control system receives the incoming trace command of input unit, image recognition is carried out;If identification is lost
It loses, return step one;Otherwise, the width d and foreign matter winding that cable diameter projects in camera plane are obtained by image recognition
The image coordinate of point;
Step 3: the actual distance Zc of foreign matter distance means is found out according to width d;
Step 4: coordinate of the laser point in the plane of delineation is calculated as coordinates of targets (x according to distance Zc0,y0);
Step 5: calculating coordinate of the plane of delineation coordinate system to angle coordinate system;Control system is according to camera under distance Zc
The corresponding relationship of pixel and angle coordinate system in plane coordinate system obtains the angle coordinate (θ of laser pointx0,θy0) and foreign matter point
Angle coordinate (θx1,θy1), by foreign matter point location;
Step 6: it is poor to make to the angle coordinate of laser point and foreign matter point, obtains angle delta θ=(Δ θ that turntable should rotatex,
Δθy), control system sends commands to turntable, is overlapped laser point and foreign matter point.
Further, for step 2, specific steps are as follows:
Step 2.1: the pixel value that control system arrives cameras capture is the image gray processing of n*m.
Step 2.2: image being converted to the form of byte array, an available n*m is (at this time for 1920*1080's)
Image array, wherein each elements AijThe gray value of the corresponding pixel.
Step 2.3: obtaining the image information for specifying region in image, be converted to image recognition matrix, matrix is indicated
Image information carries out the projection of x-axis and y-axis respectively in the coordinate system of camera plane;For every a line in image recognition matrix,
Each column seek ∑ Ai,j, curve graph is done to the result, obtains the projection in x-axis and y-axis.
Such as Fig. 8, it is assumed that be the region of specified identification in red frame.It is illustrated below with the matrix of 6*12.It is right in red frame
The gray value answered saves corresponding position in a matrix, as follows, is outlined with rectangle frame.
It is extracted, is converted to image recognition matrix:
Referring to Fig. 9, the image information that matrix (1) indicates is subjected to x-axis and y-axis respectively in the coordinate system of camera plane
Projection.
∑ A is asked for every a line, each column in matrix (1)i,j.Curve graph is made to result, it is available in x-axis and y-axis
On projection it is as shown in FIG. 10 and 11.
Step 2.4: by analyzing the obtained curve of projection, the corresponding image coordinate of available valley value, i.e. foreign matter is twined
Around a coordinate (x1,y1)。
Step 2.5: finding out in y-axis that the pixel of trough is wide, it can obtain the width that cable diameter projects in camera plane
D is spent, cable true diameter is d0Rice, and it is known.
Further, for step 3, concrete operations are as follows: similar with triangle according to the pin-hole model in camera model
Principle obtains formula:Wherein, f is camera focus, and d is the width that cable diameter projects in camera plane, d0It is line
Cable true diameter finds out the actual distance Zc of foreign matter distance means by the formula.
Pin-hole model be it is simplest in camera model, be able to satisfy the calibration request of general video camera, as shown in figure 12.
In this model, light comes from scene or far object emission, by the surface object X, passes through lens centre (aperture), In
The image x that handstand is formed in imaging plane at the distance f of lens rear, with object at handstand similarity relation.Therefore, with distant place object
The relevant image size of body is only described with a camera parameters: focal length f.
In order to facilitate consideration, projection plane is moved to it about the symmetrical position of aperture, such as Figure 13 institute on " mathematics "
Show.
According to Similar Principle of Triangle it follows that
I.e.Wherein, f is camera focus, and d is the width that cable diameter projects in camera plane, d0It is that cable is true
Real diameter, control system can be in the hope of foreign matter distance Z by the formulac。
Further, for step 4, specific steps are as follows:
Step 4.1: absolute parallel according to optical path and camera imaging axle center, it is known that phase of the laser point in camera axle center corresponding points
To fixed position, known to this relative position.In real coordinate system, if the coordinate of camera axle center corresponding points is (ac,bc), then swash
Light spot position is (ac+dx,bc+dy).Dx and dy is respectively under same distance, and laser point and camera axle center corresponding points are sat in reality
Fixation alternate position spike under mark system.
Step 4.2: the corresponding point in camera imaging axle center is in imaging process, no matter being how many with a distance from camera,
Coordinate in camera plane is constant always, is (xc,yc).If coordinate of the laser point in camera plane is (x0,y0)。(x0,y0)
=(xc+dx’,yc+dy').Dx ' and dy ' is laser point and coordinate difference of the camera axle center corresponding points under camera plane coordinate system.
Then:
Solve (x0,y0)。
Further, for step 5, specific steps are as follows:
Calculating coordinate of the plane of delineation coordinate system to angle coordinate system.Control system descends camera plane coordinate according to this distance
Pixel and the corresponding relationship of angle coordinate system obtain the angle coordinate of laser point and foreign matter point in system, by foreign matter point location.
Step 5.1: resolving laser point relative angle coordinate (θx0,θy0)。
Step 5.2: resolving foreign matter point relative angle coordinate (θx1,θy1)。
Claims (5)
1. one kind removes foreign matter device object recognition and detection method based on reflection-type laser, it is characterised in that: this method is based on anti-
-emitting laser removes foreign matter device, which includes turntable (22), laser part (4), input/output unit, control system and view
Feel sensor (10);Wherein, turntable (22) includes orientation component (1) and pitch components (2), and laser part (4) includes laser hair
Penetrate head (19), reflection mirror component one (20), reflection mirror component two (21);
Orientation component (1) is mounted on above fixed frame (5), and pitch components (2) are mounted on above orientation component (1), Laser emission
Head (19) is mounted in orientation component (1), and reflection mirror component one (20), reflection mirror component two (21) are mounted on pitch components (2)
In, visual sensor (10) is mounted in pitch components (2);
Control system is connect with visual sensor (10), adjusts visual sensor (10) parameter;Visual sensor (10) is for catching
Foreign matter point is obtained, image is transmitted to control system and carries out image recognition;Control system respectively with orientation component (1), pitch components (2)
It is connected, control system handles the result of image recognition, and driving orientation component (1), pitch components (2) carry out orientation and bow
The rotation of two-freedom is faced upward, so that visual sensor (10) and laser part (4) be driven to rotate;Input/output unit and control are
System connection, control interface and input operation information for display control program;
Method includes the following steps:
Step 1: the current image frame including foreign matter target is acquired, and is transmitted to control system, on a display screen real-time display
Image;
Step 2: after control system receives the incoming trace command of input unit, image recognition is carried out;If recognition failures,
Return step one;Otherwise, width d that cable diameter projects in camera plane and foreign matter winding point are obtained by image recognition
Image coordinate;
Step 3: the actual distance Zc of foreign matter distance means is found out according to width d;
Step 4: coordinate of the laser point in the plane of delineation is calculated as coordinates of targets (x according to distance Zc0,y0);
Step 5: calculating coordinate of the plane of delineation coordinate system to angle coordinate system;Control system is according to camera plane under distance Zc
The corresponding relationship of pixel and angle coordinate system in coordinate system obtains the angle coordinate (θ of laser pointx0,θy0) and foreign matter point angle
Spend coordinate (θx1,θy1), by foreign matter point location;
Step 6: it is poor to make to the angle coordinate of laser point and foreign matter point, obtains angle delta θ=(Δ θ that turntable should rotatex,Δ
θy), control system sends commands to turntable, is overlapped laser point and foreign matter point.
2. object recognition and detection method according to claim 1, it is characterised in that: the concrete methods of realizing of the step 2
Are as follows:
Step 2.1: the pixel value that control system arrives cameras capture is the image gray processing of n*m;
Step 2.2: image being converted to the form of byte array, obtains the image array of a n*m, wherein each elements Aij
The gray value of the corresponding pixel
Step 2.3: the image information for specifying region in image is obtained, image recognition matrix is converted to, the image that matrix is indicated
Information carries out the projection of x-axis and y-axis respectively in the coordinate system of camera plane;For every a line in image recognition matrix, each
Column seek ∑ Ai,j, curve graph is made to the result, obtains the projection in x-axis and y-axis;
Step 2.4: the curve obtained by analyzing projection obtains the corresponding image coordinate of valley value, i.e. foreign matter winding point coordinate
(x1,y1);
Step 2.5: the pixel for finding out trough in y-axis is wide to get the width d projected in camera plane to cable diameter.
3. object recognition and detection method according to claim 1, it is characterised in that: in the step 3, according to video camera
Pin-hole model and Similar Principle of Triangle in model obtain
Wherein, f is camera focus, and d is the width that cable diameter projects in camera plane, d0It is cable true diameter, by this
Formula finds out the actual distance Zc of foreign matter distance means.
4. object recognition and detection method according to claim 1, it is characterised in that: the concrete methods of realizing of the step 4
Are as follows:
Step 4.1: it is absolute parallel according to optical path and camera imaging axle center, obtain laser point in the relatively solid of camera axle center corresponding points
Place is set in positioning;In real coordinate system, if the coordinate of camera axle center corresponding points is (ac,bc), then laser spot position is (ac+dx,bc+
Dy), dx and dy is respectively laser point and fixation alternate position spike of the camera axle center corresponding points under real coordinate system under same distance;
Step 4.2: for the corresponding point in camera imaging axle center in imaging process, the coordinate in camera plane is constant always, is
(xc,yc);If coordinate of the laser point in camera plane is (x0,y0), (x0,y0)=(xc+dx’,yc+ dy '), dx ' and dy ' are
Laser point and coordinate difference of the camera axle center corresponding points under camera plane coordinate system, then:
Solve (x0,y0)。
5. object recognition and detection method according to claim 1, it is characterised in that: the concrete methods of realizing of the step 5
Are as follows:
Step 5.1: resolving laser point relative angle coordinate (θx0,θy0)
Step 5.2: resolving foreign matter point relative angle coordinate (θx1,θy1)
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810624639.5A CN110480180A (en) | 2018-06-16 | 2018-06-16 | The object recognition and detection method of foreign matter device is removed based on reflection-type laser |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810624639.5A CN110480180A (en) | 2018-06-16 | 2018-06-16 | The object recognition and detection method of foreign matter device is removed based on reflection-type laser |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110480180A true CN110480180A (en) | 2019-11-22 |
Family
ID=68545297
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810624639.5A Pending CN110480180A (en) | 2018-06-16 | 2018-06-16 | The object recognition and detection method of foreign matter device is removed based on reflection-type laser |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110480180A (en) |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR20020016440A (en) * | 2000-08-25 | 2002-03-04 | 윤종용 | System for tracking target using image feedback and Method therefor |
US20100074597A1 (en) * | 2008-09-24 | 2010-03-25 | Canon Kabushiki Kaisha | Image processing apparatus, image processing method, and program |
CN105242392A (en) * | 2015-11-16 | 2016-01-13 | 江苏省电力公司电力科学研究院 | Sighting device and method for removing foreign matter on power transmission line by means of laser remotely |
CN105655924A (en) * | 2016-03-29 | 2016-06-08 | 国网江苏省电力公司电力科学研究院 | Aiming device used for remotely removing overhead line dynamic foreign bodies through laser and method thereof |
CN206947857U (en) * | 2017-07-21 | 2018-01-30 | 山东中实电力科技有限公司 | A kind of remote power network foreign matter laser removes system |
CN108055501A (en) * | 2017-11-22 | 2018-05-18 | 天津市亚安科技有限公司 | A kind of target detection and the video monitoring system and method for tracking |
-
2018
- 2018-06-16 CN CN201810624639.5A patent/CN110480180A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR20020016440A (en) * | 2000-08-25 | 2002-03-04 | 윤종용 | System for tracking target using image feedback and Method therefor |
US20100074597A1 (en) * | 2008-09-24 | 2010-03-25 | Canon Kabushiki Kaisha | Image processing apparatus, image processing method, and program |
CN105242392A (en) * | 2015-11-16 | 2016-01-13 | 江苏省电力公司电力科学研究院 | Sighting device and method for removing foreign matter on power transmission line by means of laser remotely |
CN105655924A (en) * | 2016-03-29 | 2016-06-08 | 国网江苏省电力公司电力科学研究院 | Aiming device used for remotely removing overhead line dynamic foreign bodies through laser and method thereof |
CN206947857U (en) * | 2017-07-21 | 2018-01-30 | 山东中实电力科技有限公司 | A kind of remote power network foreign matter laser removes system |
CN108055501A (en) * | 2017-11-22 | 2018-05-18 | 天津市亚安科技有限公司 | A kind of target detection and the video monitoring system and method for tracking |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110142785A (en) | A kind of crusing robot visual servo method based on target detection | |
KR101645722B1 (en) | Unmanned aerial vehicle having Automatic Tracking and Method of the same | |
CN104067111B (en) | For following the tracks of the automated systems and methods with the difference on monitoring objective object | |
JP5081014B2 (en) | Target and three-dimensional shape measuring apparatus using the same | |
EP1462984A2 (en) | Authentication object image pick-up device and method thereof | |
WO2012020696A1 (en) | Device for processing point group position data, system for processing point group position data, method for processing point group position data and program for processing point group position data | |
CN104125372B (en) | Target photoelectric search and detection method | |
CN106441278A (en) | Unmanned aerial vehicle indoor positioning system and method | |
US20030049027A1 (en) | Automatic framing camera | |
JPH02143309A (en) | Operation method and apparatus | |
CN101057256A (en) | Face feature collating device, face feature collating method, and program | |
WO2021168838A1 (en) | Position information determining method, device, and storage medium | |
CN108154126B (en) | Iris imaging system and method | |
JP2007120993A (en) | Object shape measuring device | |
KR20170022872A (en) | Unmanned aerial vehicle having Automatic Tracking | |
CN111611989A (en) | Multi-target accurate positioning identification method based on autonomous robot | |
CN207571761U (en) | Iris imaging system | |
CN108564628A (en) | A kind of cutterhead vision positioning orientation system towards development machine automation | |
CN111780715A (en) | Visual ranging method | |
US20230219220A1 (en) | Error detection method and robot system based on association identification | |
JP5019478B2 (en) | Marker automatic registration method and system | |
CN117824624B (en) | Indoor tracking and positioning method, system and storage medium based on face recognition | |
CN110490834A (en) | The object recognition and detection method of foreign matter device is removed based on laser | |
US20200073409A1 (en) | Uav landing system and landing method thereof | |
JP2003136465A (en) | Three-dimensional position and posture decision method of detection target object and visual sensor of robot |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20191122 |
|
RJ01 | Rejection of invention patent application after publication |