CN112525109A - Method for measuring object motion attitude angle based on GPU - Google Patents
Method for measuring object motion attitude angle based on GPU Download PDFInfo
- Publication number
- CN112525109A CN112525109A CN202011510267.7A CN202011510267A CN112525109A CN 112525109 A CN112525109 A CN 112525109A CN 202011510267 A CN202011510267 A CN 202011510267A CN 112525109 A CN112525109 A CN 112525109A
- Authority
- CN
- China
- Prior art keywords
- image
- gpu
- attitude angle
- measured object
- motion attitude
- 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
-
- 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
- G01B11/26—Measuring arrangements characterised by the use of optical techniques for measuring angles or tapers; for testing the alignment of axes
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Length Measuring Devices By Optical Means (AREA)
Abstract
The invention discloses a method for measuring an object motion attitude angle based on a GPU (graphics processing Unit), which comprises the following steps: firstly, marking points are arranged on a measured object; secondly, calibrating the measured object to obtain calibration data; controlling the movement of the measured object, continuously acquiring the image of the measured object by using an industrial camera, and continuously transmitting the acquired image to an image acquisition card; fourthly, the image acquisition card transmits the image to the GPU processor, the GPU processor asynchronously extracts the mark points on the image, matches the mark points with the calibration data, and transmits the mark point data successfully matched to the host; fifthly, displaying and storing the mark point data until the object to be measured stops moving; and sixthly, sequentially carrying out preprocessing, projection, splitting, inflection point elimination, abnormal point elimination and mean value operation on all the mark point data by the host to obtain the motion attitude angle of the measured object. The invention can synchronously acquire the image of the measured object and can carry out high-speed and effective processing on the acquired image.
Description
Technical Field
The invention relates to the technical field of video measurement, in particular to a method for measuring an object motion attitude angle based on a GPU.
Background
The moving posture video measuring technology is to use a machine to replace human eyes for measurement and judgment, but the human eyes or a common industrial camera can hardly capture the details of a moving object due to the high speed of the object moving at a high speed, so that the image acquisition technology capable of imaging at a high speed is required to finish the rapid and repeated sampling of a high-speed target in a short time, so that the change process of the recorded target is clearly and slowly presented in front of the eyes of people.
The publication No. CN110849332A discloses an attitude measurement system for a moving object, which includes a controller, a driving board, a light-compensating illumination device, a temperature and humidity compensation component, and an image acquisition device, where the image acquisition device is used to acquire an image of an object to be measured and send the acquired image to the controller; the temperature and humidity compensation component is used for compensating the temperature and humidity of the image acquisition device; the light supplementing lighting device is used for compensating the illuminance of the object to be detected; the controller controls the image acquisition device, the temperature and humidity compensation assembly and the light supplementing illumination device through the driving board respectively, analyzes and processes received image information and outputs a measurement result. The technology can reduce the influence of environmental factors on measurement under complex working conditions, and accurately measure the pose parameters of the moving object. However, it cannot provide result data within 5 to 10 minutes after completion of the test, and has a technical problem of slow processing speed. Moreover, when a deformation measurement test is carried out on the continuous aircraft model, the image data of a single industrial camera at a temperature step is up to about 280GB, if the technology is still used for carrying out image data processing, at least several days are needed, and the test efficiency of the aircraft model is seriously influenced.
In addition, the technology does not consider the synchronism of the industrial camera during image acquisition, so that the acquisition precision of the image is influenced, and the precision of the post-image processing calculation is greatly influenced.
Disclosure of Invention
The invention aims to solve the problems in the prior art and provides a method for measuring the motion attitude angle of an object based on a GPU (graphics processing unit).
In order to achieve the purpose, the technical scheme adopted by the invention is as follows:
a method for measuring an object motion attitude angle based on a GPU comprises the following steps:
firstly, setting a plurality of mark points on a static measured object as required before measurement;
calibrating the measured object to obtain calibration data containing the marking points;
controlling the movement of the measured object to carry out measurement, continuously acquiring images of the measured object by using at least two industrial cameras during measurement, and continuously transmitting the acquired images to an image acquisition card;
after receiving the image, the image acquisition card transmits the image to the GPU processor through the host, and after receiving the image, the GPU processor asynchronously extracts the mark points on the image, matches the extracted mark points with the calibration data and transmits the mark point data successfully matched to the host; on the other hand, the received image is compressed, and the compressed data packet is transmitted to the host computer for storage;
after receiving the mark point data, the host computer displays the mark point data on a display on one hand and stores the mark point data in a memory on the other hand until the detected object stops moving;
and step six, sequentially carrying out preprocessing, projection, splitting, inflection point elimination, abnormal point elimination and mean value operation on all the mark point data by the host to obtain the motion attitude angle of the measured object.
In the first step, the mark points are circular rings comprising a white inner circle and a black outer circle.
In the mark points, the outer diameter of the white inner circle is 0.4-5mm, and the outer diameter of the black outer circle is 2-10 mm.
And in the second step, the measured object is calibrated by adopting an eight-step calibration method.
In the third step, the acquisition frequency of the industrial camera is 80 Hz.
In the third step, the number of the images collected by the industrial camera is 80 sheets/second.
In the third step, before the movement of the object to be measured is controlled to start measurement, the azimuth angle of the industrial camera is adjusted by the host in advance, and meanwhile, light is supplemented through an external illuminating mechanism when the light-entering quantity of the industrial camera is insufficient until the industrial camera can obtain a clear and complete image and clearly identify the mark point.
The invention has the advantages that:
1. the industrial camera can synchronously acquire the image of the measured object during measurement, and the acquisition precision of the image in the early stage and the precision of post-processing calculation are effectively improved. Meanwhile, due to the adoption of the specific GPU processor, the acquired massive images can be effectively processed at high speed, result data can be provided within 5-10 minutes after the test is finished, and the image processing efficiency is greatly improved. By adopting the specific method, the functions of real-time storage, marking point identification, sub-pixel positioning and matching and the like of massive images can be quickly and efficiently met, and a measurement result with higher accuracy can be obtained.
2. The mark points are provided with the white inner circle and the black outer circle with specific sizes, and the mark points adopting the structure have the remarkable characteristic, so that the mark points can be quickly identified and processed by a host.
3. According to the invention, the acquisition frequency of the industrial camera is set to be 80Hz, the number of images acquired by the industrial camera is set to be 80 pieces/second, the image can be acquired at high speed by the specific arrangement, and a mass of images can be obtained according to the image, so that an accurate measurement result can be obtained according to the mass of images.
4. The invention also carries out corresponding adjustment on the acquisition parameters and the like in advance before measurement, thereby being beneficial to obtaining clear acquisition images and accurate measurement results.
Detailed Description
The invention discloses a method for measuring an object motion attitude angle based on a GPU, which at least comprises two industrial cameras with adjustable azimuth angles, and is simultaneously provided with an illuminating mechanism with adjustable azimuth angles and pitch angles, and when the industrial cameras collect images, light can be supplemented through the illuminating mechanism if the illumination is insufficient. The method specifically comprises the following steps:
step one, before measurement, a measured object is clamped statically, and a plurality of mark points are arranged on the static measured object as required.
And step two, calibrating the measured object by adopting an eight-step calibration method to obtain calibration data containing the marker points, wherein the calibration data is used for later-stage matching. The shape, structure and color of the mark point are not limited, but are preferably circular rings comprising a white inner circle and a black outer circle, the outer diameter of the white inner circle is preferably 0.4-5mm, and the outer diameter of the black outer circle is preferably 2-10 mm.
And step three, controlling the movement of the measured object to start measurement, continuously acquiring the image of the measured object by using at least two industrial cameras during measurement, and continuously transmitting the acquired image to an image acquisition card. Wherein, the acquisition frequency of the industrial camera is set as 80Hz during measurement, and the number of the images acquired by the industrial camera is set as 80 pieces/second.
After receiving the image, the image acquisition card transmits the image to the GPU processor through the host, and after receiving the image, the GPU processor asynchronously extracts the mark points on the image, matches the extracted mark points with the calibration data and transmits the mark point data successfully matched to the host; and on the other hand, the received image is compressed, and the compressed data packet is transmitted to the host computer for storage.
And step five, after the host receives the mark point data, displaying the mark point data on the display on one hand, and storing the mark point data in the memory on the other hand until the detected object stops moving.
And step six, sequentially carrying out preprocessing, projection, splitting, inflection point elimination, abnormal point elimination and mean value operation on all the mark point data by the host to obtain the motion attitude angle of the measured object.
Before the movement of the object to be measured is controlled to start measurement, the method also comprises an adjusting procedure, namely the azimuth angle of the industrial camera is adjusted by the host in advance, and meanwhile, when the light inlet quantity of the industrial camera is insufficient, light is supplemented through an external illuminating mechanism until the industrial camera can obtain clear and complete images and can clearly identify the mark points.
Specifically, the adjusting process mainly comprises the following steps:
and S1, placing the object to be measured in a proper position, fixing the industrial camera horizontally, adjusting the angle position between the industrial camera and the object to be measured, turning on the illuminating mechanism to perform light compensation, completing focusing, and observing through image acquisition software in the host computer in a trial acquisition mode, wherein the standard is that the image is clear and the mark points can be normally identified.
And step S2, issuing an order to accurately adjust the angle between the industrial camera and the measured object through the host control software, simultaneously opening the image analysis processing software to observe the image acquisition imaging condition of the measured object, and finely adjusting parameters such as light, angle and the like again until clear and complete video images can be obtained in the imaging test window and all the marked points are normally and clearly identified.
In the invention, the host comprises a main control panel, the data transmission, control, processing and the like involved in the method are all completed by the main control panel, the GPU processor and the image acquisition card are integrated on the main control panel, and the display is connected with the main control panel. The specification model of the main control board can be WS X299 SAGE/10G, the specification model of the GPU processor can be GeForce RTX 2080Ti TURBO 11G, the specification model of the image acquisition card can be AS-FBD-4XCXP6-2PE8, and the specification model of the display screen can be a 17-inch capacitive touch screen.
While the invention has been described with reference to specific embodiments, any feature disclosed in this specification may be replaced by alternative features serving the same, equivalent or similar purpose, unless expressly stated otherwise; all of the disclosed features, or all of the method or process steps, may be combined in any combination, except mutually exclusive features and/or steps.
Claims (7)
1. A method for measuring an object motion attitude angle based on a GPU is characterized by comprising the following steps:
firstly, setting a plurality of mark points on a static measured object as required before measurement;
calibrating the measured object to obtain calibration data containing the marking points;
controlling the movement of the measured object to carry out measurement, continuously acquiring images of the measured object by using at least two industrial cameras during measurement, and continuously transmitting the acquired images to an image acquisition card;
after receiving the image, the image acquisition card transmits the image to the GPU processor through the host, and after receiving the image, the GPU processor asynchronously extracts the mark points on the image, matches the extracted mark points with the calibration data and transmits the mark point data successfully matched to the host; on the other hand, the received image is compressed, and the compressed data packet is transmitted to the host computer for storage;
after receiving the mark point data, the host computer displays the mark point data on a display on one hand and stores the mark point data in a memory on the other hand until the detected object stops moving;
and step six, sequentially carrying out preprocessing, projection, splitting, inflection point elimination, abnormal point elimination and mean value operation on all the mark point data by the host to obtain the motion attitude angle of the measured object.
2. The method for measuring the motion attitude angle of the object based on the GPU of claim 1, wherein: in the first step, the mark points are circular rings comprising a white inner circle and a black outer circle.
3. The method for measuring the motion attitude angle of the object based on the GPU of claim 2, wherein: in the mark points, the outer diameter of the white inner circle is 0.4-5mm, and the outer diameter of the black outer circle is 2-10 mm.
4. The method for measuring the motion attitude angle of the object based on the GPU of claim 1, wherein: and in the second step, the measured object is calibrated by adopting an eight-step calibration method.
5. The method for measuring the motion attitude angle of the object based on the GPU of claim 1, wherein: in the third step, the acquisition frequency of the industrial camera is 80 Hz.
6. The method for measuring the motion attitude angle of the object based on the GPU of claim 1, wherein: in the third step, the number of the images collected by the industrial camera is 80 sheets/second.
7. The method for measuring the motion attitude angle of the object based on the GPU of claim 1, wherein: in the third step, before the movement of the object to be measured is controlled to start measurement, the azimuth angle of the industrial camera is adjusted by the host in advance, and meanwhile, light is supplemented through an external illuminating mechanism when the light-entering quantity of the industrial camera is insufficient until the industrial camera can obtain a clear and complete image and clearly identify the mark point.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011510267.7A CN112525109A (en) | 2020-12-18 | 2020-12-18 | Method for measuring object motion attitude angle based on GPU |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011510267.7A CN112525109A (en) | 2020-12-18 | 2020-12-18 | Method for measuring object motion attitude angle based on GPU |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112525109A true CN112525109A (en) | 2021-03-19 |
Family
ID=75001673
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011510267.7A Pending CN112525109A (en) | 2020-12-18 | 2020-12-18 | Method for measuring object motion attitude angle based on GPU |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112525109A (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101825445A (en) * | 2010-05-10 | 2010-09-08 | 华中科技大学 | Three-dimension measuring system for dynamic object |
| CN102788572A (en) * | 2012-07-10 | 2012-11-21 | 中联重科股份有限公司 | Method, device and system for measuring attitude of engineering machinery lifting hook |
| US20190310084A1 (en) * | 2017-03-17 | 2019-10-10 | Dalian University Of Technology | Motion measurement method and apparatus applied to large multi-paddle wave simulation system |
| CN110796701A (en) * | 2019-10-21 | 2020-02-14 | 深圳市瑞立视多媒体科技有限公司 | Identification method, device and equipment of mark points and storage medium |
| CN110849332A (en) * | 2019-11-26 | 2020-02-28 | 成都立鑫新技术科技有限公司 | Attitude measurement system of moving object |
| CN111421539A (en) * | 2020-04-01 | 2020-07-17 | 电子科技大学 | Industrial part intelligent identification and sorting system based on computer vision |
-
2020
- 2020-12-18 CN CN202011510267.7A patent/CN112525109A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101825445A (en) * | 2010-05-10 | 2010-09-08 | 华中科技大学 | Three-dimension measuring system for dynamic object |
| CN102788572A (en) * | 2012-07-10 | 2012-11-21 | 中联重科股份有限公司 | Method, device and system for measuring attitude of engineering machinery lifting hook |
| US20190310084A1 (en) * | 2017-03-17 | 2019-10-10 | Dalian University Of Technology | Motion measurement method and apparatus applied to large multi-paddle wave simulation system |
| CN110796701A (en) * | 2019-10-21 | 2020-02-14 | 深圳市瑞立视多媒体科技有限公司 | Identification method, device and equipment of mark points and storage medium |
| CN110849332A (en) * | 2019-11-26 | 2020-02-28 | 成都立鑫新技术科技有限公司 | Attitude measurement system of moving object |
| CN111421539A (en) * | 2020-04-01 | 2020-07-17 | 电子科技大学 | Industrial part intelligent identification and sorting system based on computer vision |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109297413B (en) | Visual measurement method for large-scale cylinder structure | |
| CN109454634B (en) | Robot hand-eye calibration method based on plane image recognition | |
| CN109682575B (en) | Underwater dynamic deformation binocular measurement system for rotating blades of composite propeller model | |
| CN105405115B (en) | A kind of image mould group is to heart system and method | |
| CN107084671B (en) | A kind of recessed bulb diameter measuring system and measurement method based on three wire configuration light | |
| CN109269525B (en) | Optical measurement system and method for take-off or landing process of space probe | |
| CN108214486A (en) | control device, robot and robot system | |
| CN108188835B (en) | Machine vision-based numerical control machine tool spindle thermal elongation testing device and testing method | |
| CN112326206B (en) | AR module binocular fusion detection device and detection method | |
| CN111189401B (en) | Real-time automatic measurement method and system for shield tail clearance | |
| CN109945792A (en) | A kind of diameter measuring method, device and the application method of high reflective cylindrical body | |
| CN113269448A (en) | System and method for evaluating assembling performance of human-computer work efficiency in virtual reality environment | |
| CN109781068A (en) | The vision measurement system ground simulation assessment system and method for space-oriented application | |
| CN112082739A (en) | HUD stray light environment simulation test device and test method | |
| CN111289111A (en) | Self-calibration infrared body temperature rapid detection method and detection device | |
| CN111044262A (en) | Near-to-eye display optical-mechanical module detection device | |
| CN114155300A (en) | Projection effect detection method and device for vehicle-mounted HUD system | |
| CN109631763A (en) | Irregular part detects localization method | |
| CN112525109A (en) | Method for measuring object motion attitude angle based on GPU | |
| US4330779A (en) | Display analyzer having angular degrees of freedom | |
| CN104913721A (en) | Experimental apparatus for industrial vision sensor detection | |
| JP3777439B2 (en) | Visual inspection equipment and visual inspection method | |
| CN207894844U (en) | A kind of bearing device inside non-destructive testing imaging device | |
| CN112381735B (en) | Method for unifying AOI boundary point pixel coordinate systems of head-mounted eye tracker | |
| CN113903079B (en) | Sight line estimation method based on integrated microminiature head-mounted system |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210319 |