CN113503815A - Spraying appearance recognition method based on grating - Google Patents
Spraying appearance recognition method based on grating Download PDFInfo
- Publication number
- CN113503815A CN113503815A CN202110767058.9A CN202110767058A CN113503815A CN 113503815 A CN113503815 A CN 113503815A CN 202110767058 A CN202110767058 A CN 202110767058A CN 113503815 A CN113503815 A CN 113503815A
- Authority
- CN
- China
- Prior art keywords
- workpiece
- grating
- point cloud
- cloud data
- measuring
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
- G01B11/00—Measuring arrangements characterised by the use of optical techniques
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
- G01B7/00—Measuring arrangements characterised by the use of electric or magnetic techniques
- G01B7/02—Measuring arrangements characterised by the use of electric or magnetic techniques for measuring length, width or thickness
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Length Measuring Devices By Optical Means (AREA)
Abstract
The invention relates to spraying shape recognition, in particular to a spraying shape recognition method based on a grating. The method comprises the following steps: s1, conveying the workpiece to a workpiece measuring unit through a workpiece conveying unit; wherein the workpiece measurement unit comprises a grating measurement system and an incremental rotary encoder; wherein the workpiece conveying unit conveys the workpiece in a suspended conveying manner. S2, detecting the appearance of the workpiece through the workpiece measuring unit; s3, when the workpiece completely passes through the measuring area, the grating measuring system obtains grating point cloud data in real time, the incremental rotary encoder gives out displacement data relative to the measuring area, and the grating point cloud data and the displacement data are normalized to obtain point cloud data of the workpiece; and S4, transmitting the point cloud data of the workpiece to a system control unit for point cloud data processing to realize the appearance identification of the workpiece.
Description
Technical Field
The invention relates to spraying shape recognition, in particular to a spraying shape recognition method based on a grating.
Background
For example, publication No. 201910305106.5 discloses a method, system and storage medium for segmentation and identification based on scan point cloud data, the method includes: scanning a scene and generating a three-dimensional model of the scene from the scanned data; acquiring scanning point cloud data from a three-dimensional model of a scene; dividing the scanning point cloud data to obtain a point cloud of a first object in a scene; carrying out intelligent identification on point clouds of first objects in a scene by adopting an artificial intelligence method, wherein the intelligent identification comprises the identification of the types and the corresponding number of the first objects, and the first objects comprise indoor objects and outdoor objects; however, the identification method cannot reduce the influence of the speed control precision of the conveying chain on the workpiece modeling precision.
Disclosure of Invention
The invention provides a spraying appearance recognition method based on a grating, aiming at reducing the influence of the speed control precision of a conveying chain on the workpiece modeling precision and basically eliminating the influence of blockage on the workpiece modeling.
The above purpose is realized by the following technical scheme:
the spraying appearance recognition method based on the grating comprises the following steps:
s1, conveying the workpiece to a workpiece measuring unit through a workpiece conveying unit; wherein the workpiece measurement unit comprises a grating measurement system and an incremental rotary encoder; wherein the workpiece conveying unit conveys the workpiece in a suspended conveying manner.
S2, detecting the appearance of the workpiece through the workpiece measuring unit;
s3, when the workpiece completely passes through the measuring area, the grating measuring system obtains grating point cloud data in real time, the incremental rotary encoder gives out displacement data relative to the measuring area, and the grating point cloud data and the displacement data are normalized to obtain point cloud data of the workpiece;
and S4, transmitting the point cloud data of the workpiece to a system control unit for point cloud data processing to realize the appearance identification of the workpiece.
Wherein the grating measurement system comprises a transmitter and a receiver electrically connected to the transmitter; wherein the emitter reaches the receiver by emitting a beam, the beam not reaching the receiver in the shielded area when the workpiece is shielded between the emitter and the receiver.
The grating measuring system adopts a field bus measuring type automatic grating MLG-2ProNet, the beam distance is 10mm, and the response time is 23.3 ms; wherein the running speed of the suspension chain is 3-4 m/min.
Wherein the point cloud data processing comprises: and establishing a workpiece model according to the point cloud data of the workpiece, identifying the workpiece, estimating the pose and converting the track.
The system control unit extracts edges by using a sobel operator, acquires transverse edges and longitudinal edges respectively, and fits to obtain workpiece edge information.
Wherein the fitting employs a least squares fitting algorithm.
And calculating by adopting a RANSAC algorithm when the least square fitting algorithm fails.
The spraying appearance recognition method based on the grating has the beneficial effects that:
identifying frame type workpieces with different sizes and types by using a grating measurement system; the incremental rotary encoder is used for acquiring the displacement of the workpiece, so that the influence of the speed control precision of the conveying chain on the modeling precision of the workpiece is reduced, and the influence of the jamming on the modeling of the workpiece is basically eliminated; by processing the obtained point cloud data of the workpiece, a sample does not need to be trained, the development time is shortened, the adaptability of the product is improved, and the production requirements of multiple varieties and small batch are met; wherein the RANSAC algorithm is used to correct the workpiece tilt.
Drawings
FIG. 1 shows the relationship between identification and spraying;
FIG. 2 shows an overall process of workpiece identification and spraying;
FIG. 3 shows an example of a raster reconstructed image 1;
FIG. 4 shows an example of a raster reconstructed image 2;
FIG. 5 shows a comparison of line fitting algorithms;
FIG. 6 shows the results of workpiece analysis during point cloud data processing and tilt correction for workpiece tilt;
fig. 7 to 9 show that after gradient images in the X direction and the Y direction are obtained, edge information in the direction is obtained according to the gradient images, a linear equation is fitted by applying the RANSAC method, and an analysis result of the spraying trajectory is obtained by calculation;
fig. 10 shows the detection range of the measurement-type automated grating.
Detailed Description
With reference to fig. 1 to 10, the method for identifying the spraying shape based on the grating comprises the following steps:
s1, conveying the workpiece to a workpiece measuring unit in a suspension conveying mode through a workpiece conveying unit; wherein the workpiece measurement unit comprises a grating measurement system and an incremental rotary encoder;
the workpiece conveying unit comprises a suspension chain, and the grating measuring system adopts a field bus measuring type automatic grating MLG-2 ProNet; the measurement-type automation grating comprises a transmitter, a transmitter and a receiver, wherein the transmitter is provided with an interface for voltage supply and synchronization, and the receiver is provided with an interface for a field bus module; wherein the fieldbus module has a fieldbus interface, an ethernet interface for configuration by a PC/laptop, a power supply interface and a receiver interface, an electrical connection between the transmitter and the receiver being required. The detection range of the measurement grating is shown in fig. 10, wherein (i) represents the monitoring height, (ii) represents the beam separation, and (iii) represents the trigger sensing distance. The detection range is determined by the grating monitoring height and the scanning range. The monitoring height depends on the beam split and the number of beams. The scanning range of the grating refers to the distance between the transmitter and the receiver.
S2, detecting the workpiece through the workpiece measuring unit; the detection principle is as follows: as long as no object is located between the transmitter and the receiver, the light beam of the transmitter will reach the receiver; if an object is located between the transmitter and the receiver, the beam is interrupted according to the size of the object; the detection range is determined by the monitoring height of the grating, which depends on the beam separation and the number of beams, and the scanning range of the grating, which refers to the distance between the transmitter and the receiver. When the workpiece is carried by the suspension line to pass through the measurement grating, the interruption data of the grating is acquired in real time through the ProNet interface, and the incremental rotary encoder gives out displacement relative to the measurement area, so that the influence of the speed control precision of the conveying chain on the workpiece modeling precision is reduced, and the influence of the blockage on the workpiece modeling is basically eliminated.
S3, when the workpiece completely passes through the measuring area, the grating measuring system obtains grating point cloud data in real time, the incremental rotary encoder gives out displacement data relative to the measuring area, and the grating point cloud data and the displacement data are normalized to obtain point cloud data of the workpiece; when the workpiece completely passes through the measurement area, the workpiece is reconstructed according to the displacement data and the grating point cloud data, the normalization processing is performed to obtain the original image data of the workpiece, and the current image binary image, such as the grating reconstructed image shown in fig. 3 and 4, is obtained according to the measurement principle of the measurement grating. And (5) performing image contour analysis by using opencv, judging the number of the current detected workpieces according to the contour data, and segmenting to obtain workpiece spraying data. The method adopts the image opening operation, which is a filter based on geometric operation, different structural elements cause different filtering results, corrosion is firstly carried out and then expansion is carried out, isolated small points, burrs and small bridges can be removed, the total position and the shape are inconvenient, and the pixel precision of the edge of a workpiece is ensured.
S4, transmitting the point cloud data of the workpiece to a system control unit, namely a control computer, and processing the point cloud data; wherein the point cloud data processing: the method comprises the steps of establishing a workpiece model by point cloud data, identifying the workpiece, estimating the pose, converting the track and the like, and realizes the functions of measuring the workpieces with different types and sizes and planning the track.
S5, executing the track: and performing data conversion according to the position and attitude information of the spraying track points generated by the point cloud data and the calibration data of the robot system to obtain the motion trail of the robot to control the spraying robot, wherein the spraying robot adopts an IRB 6700 type robot, and the robot adopts an ICR5 control system.
The identification method is applied to automatic spraying of a window, a workpiece model is of a frame type, edge extraction is carried out on the system by using a sobel operator, a transverse edge and a longitudinal edge are respectively obtained, workpiece edge information is obtained through fitting, and a spraying track is generated. After extracting the transverse edge and the longitudinal edge of the workpiece, fitting to obtain the edge information of the workpiece, generally fitting by adopting a least square fitting algorithm, and influencing the fitting precision when the edge has a larger deviation point.
The function of the Soble operator integrates Gaussian smoothing and differential derivation, which is also called a first-order differential operator, and the derivation operator performs derivation in the horizontal direction and the vertical direction to obtain the gradient image of the image in the X direction and the Y direction. And acquiring edge information in the direction according to the gradient image, fitting a linear equation by using a RANSAC method, and calculating to obtain a spraying track. The analysis results are shown in fig. 7 to 9, and meet the system precision requirement.
RANSAC is an abbreviation of RANdom SAmple Consensus, and a RANdom SAmple Consensus algorithm is a probabilistic algorithm, and sometimes, in order to improve the probability of valid data, the number of iterations needs to be increased, and data is generally divided into two types: valid data (entries) and invalid data (entries), data that is not much different from the target data is valid data, and data that is much different from the target data is invalid data. If the valid data occupies most of the data, the invalid data only has a small part, and we can determine the parameters and errors of the model by the least square method or the like, if the invalid data is too much, the least square method is invalid, and a new algorithm is needed to perform, as shown in fig. 5.
The inputs to the RANSAC algorithm are a set of observations, often containing large noise or invalid points, a parameterized model for interpreting the observations, and some trusted parameters.
RANSAC achieves this goal by iteratively selecting a set of random subsets in the data. The selected subset is assumed to be an in-office point and verified by the following method:
1) there is a model adapted to the assumed local interior, i.e. all unknown parameters can be calculated from the assumed local interior.
2) All other data are tested with the model obtained in 1), and if a point is suitable for the estimated model, it is considered to be an in-office point.
3) If enough points are classified as hypothetical intra-office points, the estimated model is reasonable enough.
4) The model is then re-estimated using all hypothesized local points (e.g., using least squares), since it was estimated only by the initial hypothesized local points.
5) Finally, the model is evaluated by estimating the error rate of the local interior point and the model.
6) The above process is repeated a fixed number of times, each time the resulting model is either discarded because there are too few local points or selected because it is better than the existing models. When the suspended workpiece has an inclination, the inclination correction is performed, as shown in the lower two pictures of fig. 6, wherein the upper two pictures of fig. 6 are the workpiece analysis results.
Further, the ordinary least squares is how to achieve the optimum under the existing data. Is considered from the perspective of a minimum overall error. RANSAC is the method of first assuming that data has certain characteristics (purpose), and properly cutting off some existing data for the purpose.
Claims (10)
1. The spraying appearance recognition method based on the grating is characterized by comprising the following steps of:
s1, conveying the workpiece to a workpiece measuring unit through a workpiece conveying unit; wherein the workpiece measurement unit comprises a grating measurement system and an incremental rotary encoder;
and S2, detecting the appearance of the workpiece through the workpiece measuring unit.
2. The method of claim 1, wherein the grating measurement system comprises a transmitter and a receiver electrically connected to the transmitter; wherein the emitter reaches the receiver by emitting a beam, the beam not reaching the receiver in the shielded area when the workpiece is shielded between the emitter and the receiver.
3. The method of claim 2, wherein the grating measurement system employs a fieldbus measurement-type automated grating MLG-2 ProNet.
4. The method of claim 3, the steps further comprising:
s3, when the workpiece completely passes through the measuring area, the grating measuring system obtains grating point cloud data in real time, the incremental rotary encoder gives out displacement data relative to the measuring area, and the grating point cloud data and the displacement data are normalized to obtain point cloud data of the workpiece;
and S4, transmitting the point cloud data of the workpiece to a system control unit for point cloud data processing to realize the appearance identification of the workpiece.
5. The method of claim 4, wherein the point cloud data processing comprises: and establishing a workpiece model according to the point cloud data of the workpiece, identifying the workpiece, estimating the pose and converting the track.
6. The method of claim 5, wherein the system control unit performs edge extraction by using a sobel operator to obtain the transverse edge and the longitudinal edge respectively, and the workpiece edge information is obtained by fitting.
7. The method of claim 6, wherein the fitting employs a least squares fitting algorithm.
8. The method of claim 7, wherein the least squares fitting algorithm fails to compute using the RANSAC algorithm.
9. The method of claim 8, wherein the workpiece transport unit transports workpieces in a suspended transport manner.
10. The method according to any one of claims 3 to 6, wherein the measurement-type automated grating MLG-2ProNet has a beam distance of 10mm and a response time of 23.3 ms; the running speed of the suspension chain is 3-4 m/min.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110767058.9A CN113503815A (en) | 2021-07-07 | 2021-07-07 | Spraying appearance recognition method based on grating |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110767058.9A CN113503815A (en) | 2021-07-07 | 2021-07-07 | Spraying appearance recognition method based on grating |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113503815A true CN113503815A (en) | 2021-10-15 |
Family
ID=78011994
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110767058.9A Pending CN113503815A (en) | 2021-07-07 | 2021-07-07 | Spraying appearance recognition method based on grating |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113503815A (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2004053427A1 (en) * | 2002-12-09 | 2004-06-24 | Specialty Minerals (Michigan) Inc. | Method for positioning a measuring device emitting and receiving optical radiation for measuring wear in the lining of a container |
| WO2013096704A1 (en) * | 2011-12-20 | 2013-06-27 | Sadar 3D, Inc. | Systems, apparatus, and methods for acquisition and use of image data |
| CN205138431U (en) * | 2015-11-11 | 2016-04-06 | Tcl王牌电器(惠州)有限公司 | Material recognition device and transfer chain of transfer chain |
| CN110223297A (en) * | 2019-04-16 | 2019-09-10 | 广东康云科技有限公司 | Segmentation and recognition methods, system and storage medium based on scanning point cloud data |
| CN111307044A (en) * | 2020-04-21 | 2020-06-19 | 昆明昆船物流信息产业有限公司 | Method for measuring and calculating length, width and approximate volume of introduced material and program product |
| CN111854637A (en) * | 2020-07-14 | 2020-10-30 | 广东顺德迪峰机械有限公司 | Laser dot matrix light curtain scanning device |
-
2021
- 2021-07-07 CN CN202110767058.9A patent/CN113503815A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2004053427A1 (en) * | 2002-12-09 | 2004-06-24 | Specialty Minerals (Michigan) Inc. | Method for positioning a measuring device emitting and receiving optical radiation for measuring wear in the lining of a container |
| WO2013096704A1 (en) * | 2011-12-20 | 2013-06-27 | Sadar 3D, Inc. | Systems, apparatus, and methods for acquisition and use of image data |
| CN205138431U (en) * | 2015-11-11 | 2016-04-06 | Tcl王牌电器(惠州)有限公司 | Material recognition device and transfer chain of transfer chain |
| CN110223297A (en) * | 2019-04-16 | 2019-09-10 | 广东康云科技有限公司 | Segmentation and recognition methods, system and storage medium based on scanning point cloud data |
| CN111307044A (en) * | 2020-04-21 | 2020-06-19 | 昆明昆船物流信息产业有限公司 | Method for measuring and calculating length, width and approximate volume of introduced material and program product |
| CN111854637A (en) * | 2020-07-14 | 2020-10-30 | 广东顺德迪峰机械有限公司 | Laser dot matrix light curtain scanning device |
Non-Patent Citations (1)
| Title |
|---|
| 孙德彬: "《普通高等教育室内与家具设计专业"十三五"规划教材 家具表面装饰工艺技术 第2版》", 31 October 2018, 中国轻工业出版社 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109685066B (en) | Mine target detection and identification method based on deep convolutional neural network | |
| CN108932736B (en) | Two-dimensional laser radar point cloud data processing method and dynamic robot pose calibration method | |
| CN110497187B (en) | Sun flower pattern assembly system based on visual guidance | |
| US6470271B2 (en) | Obstacle detecting apparatus and method, and storage medium which stores program for implementing the method | |
| US20080181485A1 (en) | System and method of identifying objects | |
| CN111721259B (en) | Underwater robot recovery positioning method based on binocular vision | |
| CN109434251B (en) | Welding seam image tracking method based on particle filtering | |
| CN115592324B (en) | Automatic welding robot control system based on artificial intelligence | |
| CN112418103A (en) | Bridge crane hoisting safety anti-collision system and method based on dynamic binocular vision | |
| CN112949366B (en) | Obstacle identification method and device | |
| CN115049700A (en) | Target detection method and device | |
| CN111784655B (en) | Underwater robot recycling and positioning method | |
| CN114714355A (en) | Embedded vision tracking control system of autonomous mobile welding robot | |
| CN114488190A (en) | Laser radar 3D point cloud ground detection method | |
| CN112927303A (en) | Lane line-based automatic driving vehicle-mounted camera pose estimation method and system | |
| US20240051146A1 (en) | Autonomous solar installation using artificial intelligence | |
| CN113536959A (en) | Dynamic obstacle detection method based on stereoscopic vision | |
| CN112258517A (en) | Automatic map repairing method and device for laser radar grid map | |
| CN113503815A (en) | Spraying appearance recognition method based on grating | |
| CN113935946B (en) | Method and device for detecting underground obstacle in real time | |
| CN116678395A (en) | Navigation method and system for breeding inspection robot based on dynamic fusion of 2D laser and visual edge line | |
| CN111008607B (en) | Automatic laser scanning method and system for cabin door gap of visual servo aircraft | |
| Boby | Hand-eye calibration using a single image and robotic picking up using images lacking in contrast | |
| KR101248308B1 (en) | Method for recognizing the self position of a mobile robot unit using the lamp directionality on the ceiling image/feature map and mobile robot unit using the method | |
| CN117565062B (en) | Automatic control method and system for wafer carrying manipulator based on machine learning |
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: 20211015 |
|
| RJ01 | Rejection of invention patent application after publication |