Industrial robot performance detection method based on machine vision

Abstract

The present invention discloses an industrial robot performance detection method based on machine vision, the method comprises an infrared camera and an auxiliary imaging device, the auxiliary imaging device is installed at the front end of the infrared camera, when the product to be tested is transferred from the industrial production line to the test process outside the industrial production line, the infrared camera and the auxiliary imaging device photograph it and move it to the detection table, and then release the product to be tested to the detection table, so that the detection table outside the industrial production line can detect the product to be tested, which solves the problem in the prior art that the existing industrial production line needs to be modified to add test items, realizes the purpose of adding test items without modifying the industrial production line, and at the same time achieves the technical effect of automatic detection without manual intervention, thereby improving production efficiency.

Description

Technical Field

[0001] The present invention relates to the field of industrial robots, and in particular to a method for detecting the performance of industrial robots based on machine vision. Background Art

[0002] Industrial robots are multi-joint manipulators or multi-degree-of-freedom machine devices for the industrial field. They can perform work automatically and rely on their own power and control capabilities to achieve various functions. In the field of industrial robot production and manufacturing, it is usually necessary to test the various performances of industrial robots, especially the posture accuracy, posture repeatability, multi-directional posture accuracy change, distance accuracy, distance repeatability, position stabilization time, position overshoot, position characteristic drift, trajectory accuracy, trajectory repeatability, redirection trajectory accuracy, corner deviation, corner speed characteristics, minimum positioning time, swing deviation and other indicators specified in "GBT 12642-2013 Industrial Robot Performance Specification and Test Methods".

[0003] At present, the common measurement methods are mainly: (1) measurement methods based on three-coordinate measuring machines, such as the relevant methods in Dr. Zhang Xiaojin's degree thesis "Analysis and Modeling of Shenlian Robot Posture Accuracy" from Northeastern University; (2) measurement methods based on laser trackers, such as the academic paper "Experimental Design for Laser Tracker Measurement of Robot Distance Accuracy" published by Du Liang et al. from the National Machinery Product Safety and Quality Supervision and Inspection Center; (3) measurement methods based on wire sensors, such as the technical paper "Development of Industrial Robot Calibration System Based on Wire Sensors" published by Xiao Yongqiang et al. from Anhui Evert Intelligent Equipment Co., Ltd.; (4) measurement methods based on contact displacement sensors Methods, such as the WDL series robot detection pull-rod linear displacement sensor developed by Shanghai Tianyao Technology Co., Ltd.; (5) Measurement methods based on binocular camera machine vision, such as the scientific achievement "Research on Robot Linear Trajectory Performance Detection Technology" completed by Dong Zaili and others from the Shenyang Institute of Automation, Chinese Academy of Sciences; (6) Measurement methods based on monocular camera machine vision, such as the method of repeated positioning error caused by monocular camera machine vision measurement uncertainty proposed by Professor Mei Xuesong's team at Xi'an Jiaotong University; (7) Measurement methods based on ballbars, such as the QC10 series ballbars developed by Renishaw of the United Kingdom, which can measure and evaluate the dynamic contour accuracy of industrial robots.

[0004] Among the various schemes mentioned above, the disadvantage of scheme (1) is that it can only measure the static performance parameters of some industrial robots, the detection cost is high, and the measurement is inconvenient. The reason is that the measurement principle of the three-dimensional coordinate measuring instrument is to control the probe to contact the measured system through precise motion and detection. The measured object needs to be stationary, but it is powerless for dynamically changing objects. Therefore, it is impossible to measure the dynamic performance indicators of industrial robots; the price of the three-dimensional coordinate measuring instrument is very high; the three-dimensional coordinate measuring machine is large and difficult to move, and the industrial robot is generally not convenient to move, so it is very inconvenient to measure: the disadvantages of scheme (2) are that the detection cost is high and it is affected by the environment. The laser tracker is relatively large and the detection process is complicated. The core sensor used in this solution is expensive. The current market price is about 800,000 yuan at the low end and between 2 million and 3 million yuan at the mid-to-high end. Such costs are unaffordable for small and medium-sized enterprises and research institutes. Even if it is rented, the rental fee is more than 20,000 yuan per day: the laser tracker is sensitive to changes in ambient temperature and airflow, so it is necessary to isolate the environment during measurement, and the auxiliary work is relatively complicated; the preheating and debugging of the laser tracker are also relatively complicated, requiring professional operation. Once the process is paused, it cannot be continued and the detection needs to be restarted. The measurement work is very inconvenient: the disadvantage of solution (3) is that it can only measure the position index, but not the posture index. Another disadvantage is the low measurement accuracy. The reason is that the system is composed of multiple wire sensors, and the measurement degrees of freedom are often less than the robot degrees of freedom. Therefore, the posture index cannot be measured. Multiple wire sensors introduce installation uncertainty errors when combined for measurement, and multiple wire sensors cause error accumulation when used in series. The disadvantage of solution (4) is that it can only measure the posture repeatability index. The reason is the performance limitation of the sensor: the disadvantage of the machine vision implementation schemes of solutions (5) and (6) is the low measurement accuracy. The reason is that the motion range of industrial robots is large. As the measurement range of the visual sensor increases, the actual scale represented by each pixel will increase, so the error will increase. In addition, the large-scale measurement introduces too much interference to the visual sensor, which is not conducive to visual image processing. Large-scale measurement also makes the visual features in the target environment more complex, causing the visual sensor to work in a large dynamic performance range, which is not conducive to image stability. The disadvantage of solution (7) is that the measurement is complex and inconvenient. The reason is that the ballbar is a single-point contact measurement sensor. For dynamic measurement indicators, complex detection experiments and detection algorithms need to be designed, and the professional ability requirements of the detection personnel are relatively strict. Summary of the invention

[0005] In order to overcome the shortcomings and deficiencies of the prior art, the present invention provides an industrial robot performance detection method based on machine vision.

[0006] The technical solution adopted by the present invention is to propose an industrial robot performance detection method based on machine vision, which is applied to an industrial robot. The industrial robot includes an infrared camera and an auxiliary imaging device, and the auxiliary imaging device is installed at the front end of the infrared camera. The industrial robot performance detection method based on machine vision includes the following steps:

[0007] In response to the industrial production line quality monitoring platform rotating the industrial products with different degrees of completion to different angles, driving the auxiliary imaging device to photograph the industrial products with different degrees of completion;

[0008] After detecting that the auxiliary imaging device is shooting stably, the infrared camera is driven, and the industrial robot moves the industrial products with different degrees of completion to the preset positions of the quality control console, and the industrial production line quality monitoring platform moves;

[0009] When it is detected that the industrial products with different degrees of completion reach the preset position of the quality control console, the auxiliary imaging device is driven to release the industrial products with different degrees of completion, and a signal for starting a test is sent to the quality control console.

[0010] The beneficial effect of the present invention is that without modifying the existing industrial production line, the product to be tested is transferred from the industrial production line to a quality control console outside the industrial production line through an infrared camera and an auxiliary imaging device, and the performance test is automatically completed without manual intervention.

[0011] Based on the above technical solution, the present invention can also be improved as follows.

[0012] Preferably, when it is detected that the industrial products with different degrees of completion reach the preset position of the quality control console, the auxiliary imaging device is driven to release the industrial products with different degrees of completion, and after the step of sending a test start signal to the quality control console, the method further includes:

[0013] In response to the detection results sent by the quality control console, the infrared camera is driven, the industrial robot moves to the preset position of the quality control console, and the auxiliary imaging device is driven to photograph the industrial products with different degrees of completion;

[0014] Moving the industrial products with different degrees of completion to a product classification table or an industrial production line classifier according to the detection results;

[0015] When it is detected that the industrial products with different degrees of completion have reached qualified quality, the auxiliary imaging device is driven to release the industrial products with different degrees of completion.

[0016] The beneficial effect of adopting the above further scheme is: under the condition of manual intervention, the faulty products are automatically removed from the industrial production line according to the detection results, thereby improving the yield of the products leaving the factory, reducing the after-sales process and reducing the after-sales cost.

[0017] Based on the above technical solution, the present invention can also be improved as follows.

[0018] Preferably, in response to the detection result sent by the quality control console, the infrared camera is driven, the industrial robot moves to the preset position of the quality control console, and the auxiliary imaging device is driven to photograph the industrial products with different degrees of completion, and the method further includes:

[0019] After detecting that the auxiliary imaging device is stable in shooting, the infrared camera is driven, and the industrial robot moves the industrial products with different degrees of completion to the preset position of the performance test bench, and after waiting for the detection to be completed, the auxiliary imaging device is driven to shoot the industrial products with different degrees of completion.

[0020] The beneficial effect of adopting the above further scheme is: the present application realizes the addition of new test benches without modifying the existing industrial production lines and test benches, which provides a larger space for future upgrade needs and improves the service life of the technical solution of the present application.

[0021] Based on the above technical solution, the present invention can also be improved as follows.

[0022] Preferably, the auxiliary imaging device comprises a first image scanner and a second image scanner, the first image scanner is installed on the left side of the front end of the infrared camera, and the second image scanner is installed on the right side of the front end of the infrared camera; the industrial robot performance detection method based on machine vision also includes the following steps;

[0023] In response to the inspection result sent by the quality control console, the first image scanner is driven to photograph the product shell, and the infrared camera is driven, the industrial robot moves to the preset position of the quality control console, and the second image scanner is driven to photograph the industrial products with different degrees of completion;

[0024] Driving the infrared camera and the industrial robot to release the product shell to a preset position of the quality control console;

[0025] Moving the industrial products with different degrees of completion to a product classification table or a preset position of the performance test table according to the test results;

[0026] When it is detected that the industrial products with different degrees of completion have reached qualified quality, the auxiliary imaging device is driven to release the industrial products with different degrees of completion.

[0027] The beneficial effect of adopting the above further scheme is: this application reduces the number of times the infrared camera moves back and forth by increasing the number of image scanners of the auxiliary imaging device, improves the operating efficiency of the infrared camera in a single operation, reduces the wear and tear of the infrared camera, and increases its service life.

[0028] Based on the above technical solution, the present invention can also be improved as follows.

[0029] Preferably, the industrial robot performance detection method based on machine vision further includes the following steps:

[0030] In response to the test results sent by the quality control console and the performance test bench, the first image scanner is driven to photograph the internal components of the product, and the infrared camera is driven, the industrial robot moves to the preset position of the quality control console, the second image scanner is driven to photograph the product shell, and the internal components of the product are released to the preset position of the quality control console;

[0031] When there is no abnormal information in the detection result, the infrared camera is driven, the industrial robot moves to the preset position of the performance test bench, the first image scanner photographs industrial products with different degrees of completion, and the product shell is released to the preset position of the performance test bench;

[0032] Moving the industrial products with different degrees of completion to a product classification table or an industrial production line classifier according to the detection results;

[0033] When it is detected that the industrial products with different degrees of completion have reached qualified quality, the auxiliary imaging device is driven to release the industrial products with different degrees of completion.

[0034] The beneficial effect of adopting the above-mentioned further scheme is: the present application optimizes the workflow of two image scanners and multiple test benches, so that when the testing of the products to be tested on multiple test benches is completed, the infrared camera only needs to pass through each test bench once to replace the products to be tested on all test benches, and all products to be tested have been tested on all test benches.

[0035] Based on the above technical solution, the present invention can also be improved as follows.

[0036] Preferably, the industrial production line is equipped with a manual security inspection station; the industrial robot performance detection method based on machine vision also includes the following steps:

[0037] The detection result of the manual security inspection station is obtained, and the product to be tested is moved to a preset position of the quality control console or a product classification station according to the detection result.

[0038] The beneficial effect of adopting the above-mentioned further scheme is: this application reduces the operating time of the infrared camera and increases the service life of the infrared camera by installing a manual security inspection station that can be installed without modifying the existing industrial production line on the existing industrial production line, and can still remove the faulty parts detected by the manual security inspection station from the industrial production line without manual intervention.

[0039] Based on the above technical solution, the present invention can also be improved as follows.

[0040] Preferably, a ZigBee positioning system is installed under each product to be tested in the industrial production line, and the ZigBee positioning system has a positioning mark and corresponds one-to-one with the product to be tested; the automated detection method further includes the following steps:

[0041] Write the test results of all test benches into the corresponding ZigBee positioning system;

[0042] When the tested product is moved to the industrial production line classifier, the infrared camera is adjusted to move the tested product to the corresponding ZigBee positioning system classifier.

[0043] The beneficial effect of adopting the above further scheme is: the present application realizes that the performance of the corresponding product to be tested can be obtained by reading the test information of the ZigBee positioning system through a one-to-one correspondence between the ZigBee positioning system with a positioning mark and the product to be tested, and writes the test results of all test benches into the corresponding ZigBee positioning system, thereby facilitating the repair of faulty products and the improvement of production processes.

[0044] The present application installs a test bench outside the industrial production line and uses an industrial robot to move the products to be tested on the industrial production line to the test bench outside the industrial production line to complete the test, thereby achieving the purpose of adding test items without modifying the industrial production line, and at the same time achieving the technical effect of automatic detection without manual intervention, thereby improving production efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 This is a schematic diagram of the first process of the method for detecting performance of an industrial robot based on machine vision according to the present invention;

[0046] Figure 2 It is a schematic diagram of the second process of the method for detecting performance of an industrial robot based on machine vision of the present invention;

[0047] Figure 3 It is a schematic diagram of the third process of the method for detecting performance of an industrial robot based on machine vision of the present invention. DETAILED DESCRIPTION

[0048] The principles and features of the present invention are described below in conjunction with specific implementations. The examples given are only used to explain the present invention and are not used to limit the scope of the present invention.

[0049] Reference Figure 1 The present application proposes an industrial robot performance detection method based on machine vision, which is applied to an industrial robot. The industrial robot includes an infrared camera and an auxiliary imaging device, and the auxiliary imaging device is installed at the front end of the infrared camera. The industrial robot performance detection method based on machine vision includes the following steps:

[0050] S10, in response to the industrial production line quality monitoring platform rotating the industrial products with different degrees of completion to different angles, driving the auxiliary imaging device to photograph the industrial products with different degrees of completion;

[0051] It is easy to understand that a lifting platform is installed on the industrial production line. When the products to be tested flow to the lifting platform along the conveyor belt, the lifting platform will rotate industrial products with different degrees of completion to different angles, preventing the industrial robot from touching other equipment on the industrial production line during shooting, thereby reducing the working failure rate of the industrial production line and industrial robots.

[0052] S20, after detecting that the auxiliary imaging device is stable in shooting, the infrared camera is driven, the industrial robot moves the industrial products with different degrees of completion to the preset positions of the quality control console, and the industrial production line quality monitoring platform moves;

[0053] It should be noted that since the appearance of the products to be tested is relatively similar and the shooting position is relatively fixed, in the actual production process, it is only necessary to debug the first shooting and then apply it to other products to be tested according to the same shooting parameters. The technical solution of the present application can also determine the stability of the shooting release by installing sensors. By installing sensors, it is possible to shoot any product to be tested without debugging.

[0054] S30. When it is detected that the industrial products with different degrees of completion reach the preset position of the quality control console, the auxiliary imaging device is driven to release the industrial products with different degrees of completion, and the quality control console detects the industrial products with different degrees of completion.

[0055] It is worth emphasizing that since the test bench has high requirements on the placement of the product to be tested during the testing process, the preset position of the quality control console is a standard position at which the quality control console can directly start testing after the product to be tested is placed on the quality control console; in this embodiment, when the position of the product to be tested is correctly placed, the industrial robot starts the quality control console to realize the detection of industrial products with different degrees of completion. At the same time, by setting a trigger switch on the test bench, the test bench can automatically start detection without the industrial robot starting the quality control console.

[0056] This application uses an industrial robot to photograph the product to be tested, and uses the high precision of the industrial robot to place it on a test bench outside the industrial production line where testing can be started directly, thereby achieving automatic performance testing without modifying the industrial production line and without human interference.

[0057] Reference Figure 2 , when it is detected that the industrial products with different degrees of completion reach the preset position of the quality control console, the auxiliary imaging device is driven to release the industrial products with different degrees of completion, and after the quality control console detects the industrial products with different degrees of completion, it also includes:

[0058] S40, in response to the detection result sent by the quality control console, driving the infrared camera, the industrial robot moves to the preset position of the quality control console, and driving the auxiliary imaging device to photograph the industrial products with different degrees of completion;

[0059] It is easy to understand that when the test results sent by the quality control console are received, it means that the performance test of the quality control console has been completed, and the tested products need to be moved out of the test bench to facilitate the testing of the next product. Therefore, the infrared camera and auxiliary imaging device are driven to capture industrial products with different degrees of completion.

[0060] S50, moving the industrial products with different degrees of completion to a product classification table or an industrial production line classifier according to the detection results;

[0061] It should be noted that since there are certain performance requirements for the products to be tested, those that meet the performance requirements are qualified products, and those that do not meet the performance requirements are faulty products. Therefore, after reading the test results, they are compared with the target performance requirements to determine whether the product is a faulty part, and then decide whether to move it to the industrial production line for the next process or move it to the product classification table to wait for repair and analysis.

[0062] S60: When it is detected that the industrial products with different degrees of completion have reached qualified quality, the auxiliary imaging device is driven to release the industrial products with different degrees of completion.

[0063] It is worth emphasizing that the qualified quality will be different according to the test results of industrial products with different degrees of completion and the test order of industrial products with different degrees of completion. In the present embodiment, if the test result is a faulty part, the qualified quality is the classifier of the product classification table; if the test result is a qualified product, the qualified quality is the sequential position corresponding to the industrial products with different degrees of completion when they leave the industrial production line. For example, when leaving the industrial production line, if the industrial product with different degrees of completion is the front one of all products to be tested, the qualified quality is the front of all products to be tested; if the industrial product with different degrees of completion is the second one from the front to the back of all products to be tested, the qualified quality is after the first product to be tested and before the second product to be tested from the front to the back of all products to be tested.

[0064] This embodiment separates faulty parts from industrial production lines by analyzing test results, thereby reducing after-sales processes and costs and improving product quality. By analyzing and repairing faulty parts, it reduces the waste of raw materials and is also helpful to a certain extent in analyzing the causes of faults, optimizing production processes and improving production efficiency.

[0065] Reference Figure 3 , in response to the detection result sent by the quality control console, driving the infrared camera, the industrial robot moves to the preset position of the quality control console, and driving the auxiliary imaging device to photograph the industrial products with different degrees of completion, further comprising:

[0066] S41. After detecting that the auxiliary imaging device is stable in shooting, the infrared camera is driven, and the industrial robot moves the industrial products with different degrees of completion to a preset position on the performance test bench, and after waiting for the detection to be completed, the auxiliary imaging device is driven to shoot the industrial products with different degrees of completion.

[0067] It is easy to understand that the technical solution of this embodiment only requires the addition of a test bench so that after one test of the product to be tested is completed, another test can be performed sequentially without having to modify the devices already installed inside or outside the industrial production line.

[0068] Specifically, the auxiliary imaging device includes a first image scanner and a second image scanner, the first image scanner is installed on the left side of the front end of the infrared camera, and the second image scanner is installed on the right side of the front end of the infrared camera; the industrial robot performance detection method based on machine vision also includes the following steps;

[0069] In response to the inspection result sent by the quality control console, the first image scanner is driven to photograph the product shell, and the infrared camera is driven, the industrial robot moves to the preset position of the quality control console, and the second image scanner is driven to photograph the industrial products with different degrees of completion;

[0070] It should be noted that the present application uses two image scanners to form an auxiliary imaging device, so that the infrared camera only needs to move once to photograph the tested product and place a new product to be tested for testing, thereby improving the operating efficiency of the product to be tested from the industrial production line to the test bench, reducing the use loss of the infrared camera, and increasing the service life of the industrial robot.

[0071] Drive the infrared camera and the position of the industrial robot to release the product shell to the preset position of the quality control console;

[0072] It is worth noting that after the tested products are photographed, the test bench is empty. At this time, the untested products on the second image scanner are released to the test bench for testing, which reduces the idle time of the test bench and improves the test efficiency.

[0073] Moving the industrial products with different degrees of completion to a product classification table or a preset position of the performance test table according to the test results;

[0074] It should be emphasized that if after a test is completed, it is known that industrial products with different degrees of completion are unqualified, they can be directly moved to the product classification table to reduce the test time wasted on unqualified products, improve test efficiency, and increase production capacity.

[0075] When it is detected that the industrial products with different degrees of completion have reached qualified quality, the auxiliary imaging device is driven to release the industrial products with different degrees of completion.

[0076] It is easy to understand that in this embodiment, when industrial products with different degrees of completion pass the test at the quality control console, the qualified quality is the test position of the performance test bench, and when industrial products with different degrees of completion fail the test at the quality control console, the qualified quality is the product classification bench.

[0077] Specifically, the industrial robot performance detection method based on machine vision also includes the following steps:

[0078] In response to the test results sent by the quality control console and the performance test bench, the first image scanner is driven to photograph the internal components of the product, and the infrared camera is driven, the industrial robot moves to the preset position of the quality control console, the second image scanner is driven to photograph the product shell, and the internal components of the product are released to the preset position of the quality control console;

[0079] It should be noted that when two test benches report test results at the same time, this embodiment reduces the idle time of the test benches through two image scanners, improves the test efficiency of the test benches, and improves the technical solution.

[0080] When there is no abnormal information in the detection result, the infrared camera is driven, the industrial robot moves to the preset position of the performance test bench, the first image scanner photographs industrial products with different degrees of completion, and the product shell is released to the preset position of the performance test bench;

[0081] It is worth emphasizing that when the product to be tested passes the test, it is moved to the performance test bench for the next test, and another image scanner takes a picture of the product to be tested that has completed the test. In this way, the operation is cyclically performed in sequence so that all products that have passed the test can be tested by all test benches.

[0082] Moving the industrial products with different degrees of completion to a product classification table or an industrial production line classifier according to the detection results;

[0083] It is worth noting that, through the above-mentioned testing method, the products finally moved to the industrial production line classifier have all been tested by all the test benches, while the faulty parts that have not completed the tests of all the test benches are all moved to the product classification table, waiting for subsequent maintenance and analysis.

[0084] When it is detected that the industrial products with different degrees of completion have reached qualified quality, the auxiliary imaging device is driven to release the industrial products with different degrees of completion.

[0085] It should be emphasized that due to the heavy weight of the product to be tested, improper operation may damage the industrial production line or product classification table. Therefore, the height of the qualified quality setting is farther than the industrial production line or product classification table, and the process of releasing the product to be tested is usually set to be released when the product to be tested just touches the industrial production line or product classification table.

[0086] The embodiment itself improves the test efficiency of multiple test benches and multiple test products by improving the test methods of multiple test benches and multiple test products, reduces the idle time of each test bench, and improves the production capacity.

[0087] Specifically, the industrial production line is equipped with a manual security inspection station; the industrial robot performance detection method based on machine vision also includes the following steps:

[0088] The detection result of the manual security inspection station is obtained, and the product to be tested is moved to a preset position of the quality control console or a product classification station according to the detection result.

[0089] It is easy to understand that not all test benches cannot be installed without modifying the industrial production line. When the test bench that needs to be added can be installed without modifying the industrial production line, it is installed in the industrial production line, and the faulty parts that fail the test are moved to the product classification table by industrial robots, and the products to be tested that pass the test are moved to other test benches. Automated testing can also be achieved, which improves the efficiency and accuracy of the test, increases production capacity, and improves product quality.

[0090] Specifically, a ZigBee positioning system is installed under each product to be tested in the industrial production line, and the ZigBee positioning system has a positioning mark and corresponds one-to-one with the product to be tested; the automatic detection method also includes the following steps:

[0091] Write the test results of all test benches into the corresponding ZigBee positioning system;

[0092] It should be noted that, in this embodiment, the ZigBee positioning system not only reduces the damage to the conveyor belts of the industrial production line, but also records the test information of the products to be tested, so that after the production is completed, the products can be classified into different grades according to different test results, and the cause of the failure of the faulty parts can be located, which improves the production process to a certain extent and improves the product quality.

[0093] When the tested product is moved to the industrial production line classifier, the infrared camera is adjusted to move the tested product to the corresponding ZigBee positioning system classifier.

[0094] Since the ZigBee positioning system corresponds one-to-one with the product to be tested, the positioning mark information in the ZigBee positioning system cannot be read without tools. Therefore, in order to intuitively see the correspondence between the ZigBee positioning system and the product to be tested, during the production process, only the corresponding product to be tested is placed on the ZigBee positioning system to prevent engineers and workers from making mistakes in the test results of the products.

Claims

1. An industrial robot performance detection method based on machine vision, characterized in that: Applied to industrial robots, the industrial robots include infrared cameras and auxiliary imaging devices, the auxiliary imaging devices are installed at the front end of the infrared cameras, and the performance detection method of industrial robots based on machine vision includes the following steps: In response to the industrial production line quality monitoring platform rotating industrial products with different degrees of completion to different angles, the auxiliary imaging device is driven to shoot the industrial products with different degrees of completion; after detecting that the auxiliary imaging device is shooting stably, the infrared camera is driven to move the industrial products with different degrees of completion to the preset positions of the quality control console, and the movement of the industrial production line quality monitoring platform is controlled; The auxiliary imaging device includes a first image scanner and a second image scanner, the first image scanner is installed on the left side of the front end of the infrared camera, and the second image scanner is installed on the right side of the front end of the infrared camera; The machine vision-based industrial robot performance detection method also includes the following steps: in response to the detection results sent by the quality control console and the performance test bench, driving the first image scanner to photograph the internal components of the product, and driving the infrared camera to move the industrial products with different degrees of completion to the preset position of the quality control console, driving the second image scanner to photograph the product shell, and releasing the internal components of the product to the preset position of the quality control console; when there is no abnormal information in the detection results, driving the infrared camera to move the industrial products with different degrees of completion to the preset position of the performance test bench, the first image scanner to photograph the industrial products with different degrees of completion, and releasing the product shell to the preset position of the performance test bench; moving the industrial products with different degrees of completion to the product classification table or the industrial production line classifier according to the detection results; when it is detected that the industrial products with different degrees of completion have reached qualified quality, driving the auxiliary imaging device to release the industrial products with different degrees of completion; The industrial production line is equipped with a manual security inspection station; the industrial robot performance detection method based on machine vision also includes the following steps: obtaining the detection result of the manual security inspection station, and moving the product to be tested to a preset position of the quality control console or a product classification station according to the detection result; A ZigBee positioning system is installed under each product to be tested in the industrial production line. The ZigBee positioning system has a positioning mark and corresponds one-to-one with the product to be tested. The detection method also includes the following steps: writing the test results of all test benches into the corresponding ZigBee positioning system; when the tested product to be tested is moved to the industrial production line classifier, adjusting the infrared camera and moving the tested product to be tested to the corresponding ZigBee positioning system classifier.

Images