An AI-based vision-based intelligent detection data management platform for air conditioner appearance

Abstract

The present invention relates to the technical field of intelligent detection data management of air conditioner appearance, and specifically discloses an intelligent detection data management platform for air conditioner appearance based on AI vision, the platform comprising: a position correction evaluation module, a code reader scanning module, an air conditioner appearance quality inspection module, a database and a detection accuracy evaluation module; the present invention corrects the air conditioner product whose position is not corrected through the air valve body inside the quality inspection equipment, and evaluates whether the appearance quality inspection of the air conditioner product is qualified according to the quality inspection data corresponding to each component in each detection group in each detection surface of the air conditioner product, and finally performs manual spot check on the air conditioner products with qualified appearance quality inspection, and conducts in-depth analysis on the accuracy of detection work of the detection post according to the manual spot check data, thereby reducing the complexity of quality inspection, shortening the detection time, and improving the timeliness of discovering errors or omissions in the quality inspection process, so as to help enterprises accurately understand the actual effect and existing problems of the quality inspection link.

Description

Technical Field

[0001] The present invention relates to the technical field of air conditioner appearance intelligent detection data management technology, and in particular to an air conditioner appearance intelligent detection data management platform based on AI vision. Background Art

[0002] With the advancement of technology and the improvement of people's living standards, air conditioners, as indispensable electrical equipment in homes and commercial places, are increasingly in demand year by year. However, the appearance quality of air conditioners directly affects consumers' willingness to buy and the market competitiveness of products. Therefore, accurate and efficient inspection of air conditioner appearance has become a vital part of the air conditioner manufacturing industry.

[0003] The following problems still exist in the prior art: 1. When the air-conditioning product arrives at the inspection site, it is currently mainly dependent on the inspectors to manually adjust its position, without the automatic adjustment of the position by the air valve body inside the quality inspection equipment. The manual adjustment by the inspectors wastes a lot of time and energy, reducing the inspection efficiency of the appearance quality inspection of the air-conditioning product. In addition, multiple adjustments may also increase the risk of human error.

[0004] 2. In the appearance quality inspection of air-conditioning products, the quality inspection categories required for each inspection surface of the air-conditioning products were not classified, which increased the complexity of the quality inspection and led to extended inspection time.

[0005] 3. Currently, after the appearance quality inspection of air-conditioning products, the products are directly sealed offline. No manual spot check is carried out on the air-conditioning products that have passed the appearance quality inspection. The accuracy of the inspection work of the inspection posts is not deeply analyzed based on the manual spot check data. The accuracy and reliability of the inspection posts cannot be verified, and errors or omissions in the quality inspection process cannot be discovered in time, which increases the risk of product quality problems. At the same time, it makes it impossible for enterprises to accurately understand the actual effects and existing problems of the quality inspection link. Summary of the invention

[0006] In view of this, in order to solve the problems raised in the above background technology, an air conditioner appearance intelligent detection data management platform based on AI vision is proposed.

[0007] The object of the present invention can be achieved through the following technical solutions: The present invention provides an air conditioner appearance intelligent detection data management platform based on AI vision, including: a position correction evaluation module, which is used to guide each air conditioner product in the current batch to the detection position of the target production line through the track in turn, construct a three-dimensional information model of each air conditioner product in the current batch in the detection position, obtain the coordinate value of each feature point on each air conditioner product in the current batch, and evaluate whether the position of each air conditioner product in the current batch in the detection position is corrected. If the position is corrected, the code reader scanning module is executed, and if the position is not corrected, the position correction operation is performed;

[0008] The barcode reader scanning module is used to scan the serial numbers of the air conditioner products in the current batch after the positions of the products are corrected. If the automatic scanning fails, an audible and visual alarm is triggered to guide manual scanning for confirmation. If the automatic scanning succeeds, the air conditioner appearance quality inspection module is executed.

[0009] The air conditioner appearance quality inspection module is used to collect the quality inspection images corresponding to each inspection surface of each air conditioner product in the current batch, obtain the quality inspection data corresponding to each component in each inspection group in each inspection surface of each air conditioner product from the quality inspection images, and evaluate whether the appearance quality inspection of each air conditioner product in the current batch is qualified. If the appearance quality inspection is qualified, the system displays PASS; if the appearance quality inspection is unqualified, the system displays FAIL and a red light and buzzer alarm sound;

[0010] The database is used to store the reference positions corresponding to the characteristic points of the air-conditioning products in the quality inspection equipment, and to store the standard positions of the monitoring points corresponding to the components in the various inspection groups in the inspection surfaces of the air-conditioning products.

[0011] The detection accuracy assessment module is used to record air-conditioning products that have passed the appearance quality inspection as qualified products. Quality inspectors conduct random inspections on qualified products in the current batch, extract the number of unqualified products found in the random inspections, and extract the manual random inspection data corresponding to each historical batch of the target production line to evaluate whether the inspection work of the inspection post in the target production line is accurate. If the inspection work is inaccurate, feedback will be provided immediately.

[0012] Compared with the prior art, the embodiments of the present invention have at least the following advantages or beneficial effects: (1) The present invention evaluates whether the position of each air-conditioning product in the current batch is corrected in the inspection post. If the position is not corrected, the position is automatically corrected through the air valve body inside the quality inspection equipment, thereby improving the inspection efficiency of the appearance quality inspection of the air-conditioning product and avoiding multiple adjustments by the inspection personnel, which increases the error in the acquisition of the appearance data of the air-conditioning product.

[0013] (2) The present invention evaluates whether the appearance quality inspection of each air-conditioning product in the current batch is qualified based on the quality inspection data corresponding to each component in each inspection group in each inspection surface of each air-conditioning product, thereby reducing the complexity of quality inspection, shortening the inspection time, and improving the standardization of quality inspection work.

[0014] (3) The present invention can help verify the accuracy and reliability of the inspection post by manually spot-checking the air-conditioning products that have passed the appearance quality inspection, and conduct in-depth analysis of the inspection work accuracy of the inspection post based on the manual spot-check data, thereby improving the timeliness of discovering errors or omissions in the quality inspection process, reducing the risk of product quality problems, and helping enterprises to accurately understand the actual effects and existing problems of the quality inspection process. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for describing the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0016] Figure 1 It is a schematic diagram of the system module structure connection of the present invention.

[0017] Figure 2 This is a flow chart of the intelligent detection of the air conditioner appearance according to the present invention. DETAILED DESCRIPTION

[0018] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0019] See also Figure 1 and Figure 2 As shown, the present invention provides an air conditioner appearance intelligent detection data management platform based on AI vision, including: a position correction evaluation module, a code reader scanning module, an air conditioner appearance quality inspection module, a database and a detection accuracy evaluation module.

[0020] The position correction evaluation module is connected to the code reader scanning module, the code reader scanning module is connected to the air conditioner appearance quality inspection module, and the air conditioner appearance quality inspection module is connected to the database, the detection accuracy evaluation module and the position correction evaluation module.

[0021] The position correction evaluation module is used to guide each air-conditioning product in the current batch to the inspection position of the target production line through the track in sequence, construct a three-dimensional information model of each air-conditioning product in the current batch at the inspection position, obtain the coordinate value of each feature point on each air-conditioning product in the current batch, and evaluate whether the position of each air-conditioning product in the current batch at the inspection position is corrected. If the position is corrected, the code reader scanning module is executed; if the position is not corrected, the position correction operation is performed;

[0022] In a specific embodiment of the present invention, the specific process of evaluating whether the position of each air-conditioning product in the current batch in the inspection position is corrected is as follows: A1, extracting the reference position corresponding to each feature point of the air-conditioning product in the quality inspection equipment from the database, and bringing it into the three-dimensional information model, and obtaining the reference position coordinates corresponding to each feature point of the air-conditioning product in the quality inspection equipment, recorded as ,in, Indicates the number of the feature point. ;

[0023] A2. Record the coordinate values ​​of each feature point on each air conditioner product in the current batch as ,in, Indicates the product number of the air conditioner. ;

[0024] A3. Calculate the position offset index of each feature point on each air conditioner product in the current batch , ,in, and Respectively indicate the setting permission axis, Axis and The position offset value corresponding to the axis;

[0025] A4. Calculate the position deviation index of each air conditioner product in the current batch , ,in, Indicates the number of feature points;

[0026] A5. Compare the position offset index of each air-conditioning product in the current batch with the position offset index of the set reference. If the position offset index of an air-conditioning product in the current batch is smaller than the position offset index of the set reference, it indicates that the position of the air-conditioning product in the current batch at the inspection post has been corrected. Otherwise, it indicates that the position of the air-conditioning product in the current batch at the inspection post has not been corrected.

[0027] In a specific embodiment of the present invention, the specific method of performing the position correction operation is: by adjusting the air pressure of the valve body inside the quality inspection equipment, the push rod is driven to move precisely. When the position of the air-conditioning product moves, the coordinate values ​​of each feature point on the air-conditioning product are collected again, and the position of the air-conditioning product is evaluated again according to the evaluation method of whether the position of the air-conditioning product in the inspection position is corrected. If the position is corrected, the position correction operation is ended. If the position is still not corrected, the position correction operation is continued, and so on, until the position of the air-conditioning product is corrected.

[0028] The embodiment of the present invention evaluates whether the position of each air-conditioning product in the current batch in the inspection post is correct. If the position is not correct, the position is automatically corrected through the air valve body inside the quality inspection equipment, thereby improving the inspection efficiency of the appearance quality inspection of the air-conditioning product and avoiding multiple adjustments by the inspectors which increase the error in the acquisition of the appearance data of the air-conditioning product.

[0029] The code reader scanning module is used to start scanning to obtain the serial number of each air-conditioning product after the position of each air-conditioning product in the current batch is corrected. If the automatic scanning fails, an audible and visual alarm is triggered to guide manual scanning for confirmation. If the automatic scanning succeeds, the air-conditioning appearance quality inspection module is executed;

[0030] The air conditioner appearance quality inspection module is used to collect the quality inspection images corresponding to each inspection surface of each air conditioner product in the current batch, obtain the quality inspection data corresponding to each component in each inspection group in each inspection surface of each air conditioner product from the quality inspection images, and evaluate whether the appearance quality inspection of each air conditioner product in the current batch is qualified. If the appearance quality inspection is qualified, the system displays PASS; if the appearance quality inspection is unqualified, the system displays FAIL and a red light and buzzer alarm are on;

[0031] It should be noted that the quality inspection images corresponding to the inspection surfaces of the air-conditioning products are acquired by high-definition cameras installed on various surfaces inside the quality inspection equipment.

[0032] In a specific embodiment of the present invention, various detection groups on the top surface of the air-conditioning product include but are not limited to top cover detection, various detection groups on the front surface include but are not limited to LOGO detection, air grille detection and screw detection, various detection groups on the back surface include but are not limited to screw detection and ambient temperature sensor detection, various detection groups on the left surface include but are not limited to screw detection, and various detection groups on the right surface include but are not limited to valve body and bonnet detection, drain elbow detection and screw detection.

[0033] In a specific embodiment of the present invention, the quality inspection data includes the location of each monitoring point, the number of external damaged parts and the area of ​​each external damaged part.

[0034] It should be noted that the number of appearance damaged parts and the area of ​​each appearance damaged part are obtained by importing the collected quality inspection images into professional image processing software and processing them. This technology is mature and will not be described in detail here.

[0035] In a specific embodiment of the present invention, the specific process of evaluating whether the appearance quality inspection of each air-conditioning product in the current batch is qualified is as follows: B1, extracting the position of each monitoring point, the number of appearance damages and the area of ​​each appearance damage from the quality inspection data corresponding to each component in each inspection group in each inspection surface of each air-conditioning product, and calculating the appearance quality inspection qualification corresponding to each component in each inspection group in each inspection surface of each air-conditioning product ,in, Indicates the number of the detection surface. , Indicates the number of each test group. , Indicates the component number, ;

[0036] In a specific embodiment of the present invention, the specific process of calculating the appearance quality inspection qualification of each component in each detection group in each detection surface of each air-conditioning product is as follows: C1, based on the positions of each monitoring point corresponding to each component in each detection group in each detection surface of each air-conditioning product, the number of appearance damages and the area of ​​each appearance damage, respectively calculate the placement position qualification of each component in each detection group in each detection surface of each air-conditioning product and appearance completeness ;

[0037] It should be noted that the specific process of calculating the qualification of the placement position corresponding to each component in each detection group in each detection surface of each air-conditioning product is as follows: the positions of each monitoring point corresponding to each component in each detection group in each detection surface of each air-conditioning product are brought into the three-dimensional information model of each air-conditioning product in the detection position, and the position coordinates of each monitoring point corresponding to each component in each detection group in each detection surface of each air-conditioning product are obtained, and recorded as ,in, Indicates the number of the monitoring point. .

[0038] The standard positions of each monitoring point corresponding to each component in each detection group in each detection surface of the air-conditioning product are extracted from the database and brought into the three-dimensional information model to obtain the standard position coordinates of each monitoring point corresponding to each component in each detection group in each detection surface of the air-conditioning product and record them as .

[0039] Calculate the placement qualification of each component in each test group on each test surface of each air-conditioning product , ,in, Indicates the number of monitoring points.

[0040] It should be noted that the specific process of calculating the appearance integrity of each component in each test group on each test surface of each air-conditioning product is as follows: the number of appearance damages corresponding to each component in each test group on each test surface of each air-conditioning product is recorded as .

[0041] The damaged areas of each component in each test group on each test surface of each air conditioner product are accumulated to obtain the damaged areas of each component in each test group on each test surface of each air conditioner product. .

[0042] Calculate the appearance integrity of each component in each test group on each test surface of each air-conditioning product , ,in, and They respectively represent the number of external damages and the external damage area of ​​the set reference.

[0043] C2. Calculate the appearance quality inspection pass of each component in each test group on each test surface of each air-conditioning product , ,in, and They represent the weights of the set placement position qualification and appearance integrity corresponding to the appearance quality inspection qualification assessment, .

[0044] In a specific embodiment of the present invention, The setting value of is 0.5. The value of is set to 0.5. When calculating the appearance quality inspection qualification of air conditioner products, the installation position qualification and appearance integrity are both indispensable evaluation indicators. Their respective importance depends on the specific evaluation objectives and product characteristics, and they need to be comprehensively considered and the weights and scoring standards need to be reasonably set.

[0045] B2. Calculate the appearance quality inspection pass rate corresponding to each test group in each test surface of each air-conditioning product ;

[0046] In a specific embodiment of the present invention, the specific process of calculating the appearance quality inspection pass corresponding to each type of inspection group in each inspection surface of each air-conditioning product is as follows: D1. Compare the appearance quality inspection pass corresponding to each component in each type of inspection group in each inspection surface of each air-conditioning product with the appearance quality inspection pass of a set reference. If the appearance quality inspection pass corresponding to a component in a certain type of inspection group in a certain inspection surface of an air-conditioning product is less than the appearance quality inspection pass of the set reference, then the component in the inspection group in the inspection surface of the air-conditioning product is recorded as an abnormal component, and the number of abnormal components in each type of inspection group in each inspection surface of each air-conditioning product is counted and recorded as ;

[0047] D2. Calculate the appearance quality inspection pass rate corresponding to each test group in each test surface of each air-conditioning product , ,in, Indicates the number of components, Indicates the ratio of abnormal components to the set reference.

[0048] B3. Extract the minimum value from the appearance quality inspection pass degree corresponding to each test group in each test surface of each air-conditioning product as the appearance quality inspection pass degree corresponding to each test surface of each air-conditioning product, and compare it with the appearance quality inspection pass degree of the set reference test surface. If the appearance quality inspection pass degree corresponding to any test surface of an air-conditioning product is less than the appearance quality inspection pass degree of the set reference test surface, it indicates that the appearance quality inspection of the air-conditioning product has failed. Otherwise, it indicates that the appearance quality inspection of the air-conditioning product has passed.

[0049] The embodiment of the present invention evaluates whether the appearance quality inspection of each air-conditioning product in the current batch is qualified according to the quality inspection data corresponding to each component in each inspection group in each inspection surface of each air-conditioning product, thereby reducing the complexity of quality inspection, shortening the inspection time, and improving the standardization of quality inspection work.

[0050] The database is used to store the reference positions of each feature point of the air-conditioning product in the quality inspection equipment, and store the standard positions of each monitoring point corresponding to each component in each inspection group in each inspection surface of the air-conditioning product. The data sources in the database of this embodiment are shown in Table 1 below.

[0051] Table 1 Data sources in the database

[0052]

[0053] The detection accuracy assessment module is used to record air-conditioning products that have passed the appearance quality inspection as qualified products. Quality inspectors conduct random inspections on qualified products in the current batch, extract the number of unqualified products found in the random inspections, and extract the manual random inspection data corresponding to each historical batch of the target production line to evaluate whether the inspection work of the inspection post in the target production line is accurate. If the inspection work is inaccurate, feedback is provided immediately.

[0054] It should be noted that the number of unqualified products sampled, as well as the number of qualified air-conditioning products corresponding to each historical batch mentioned later and the number of unqualified air-conditioning products sampled are all extracted from the manual sampling data registration manual.

[0055] In a specific embodiment of the present invention, the manual inspection data includes the number of qualified air-conditioning products and the number of unqualified air-conditioning products inspected.

[0056] In a specific embodiment of the present invention, the specific process of evaluating whether the detection work of the detection post in the target production line is accurate is as follows: E1. Count the number of qualified products in the current batch and record it as ;

[0057] E2. Record the number of unqualified products sampled as ;

[0058] E3. Calculate the accuracy of the testing work of the testing position in the current batch testing , ,in, represents a natural constant, Indicates the percentage of substandard air-conditioning products used as reference;

[0059] E4. Extract the number of qualified air-conditioning products and the number of unqualified air-conditioning products from the manual sampling data corresponding to each batch in the history of the target production line. Calculate the accuracy of the inspection work of the inspection post in the inspection of each batch in the history in the same way as the calculation method of the inspection work accuracy of the inspection post in the current batch inspection. ,in, Indicates the number of each historical batch, ;

[0060] E5. Calculate the accuracy of the inspection work at the inspection post in the target production line ;

[0061] In a specific embodiment of the present invention, the specific process of calculating the detection work accuracy of the detection post in the target production line is as follows: F1, extracting the maximum value and the minimum value from the detection work accuracy of the detection post in each batch of historical detections, and recording them as and ;

[0062] F2. Calculate the accuracy of the inspection work at the inspection post in the target production line , ,in, It means that the testing position is the first in history The accuracy of the testing work in batch testing, and They represent the detection accuracy deviation and the detection accuracy extreme value difference of the set reference respectively. Indicates the number of historical batches.

[0063] E6. Compare the detection work accuracy of the detection post in the target production line with the detection work accuracy of the set reference. If the detection work accuracy of the detection post in the target production line is greater than or equal to the detection work accuracy of the set reference, it indicates that the detection work of the detection post in the target production line is accurate. Otherwise, it indicates that the detection work of the detection post in the target production line is inaccurate.

[0064] The embodiment of the present invention can help verify the accuracy and reliability of the inspection post by manually spot-checking the air-conditioning products that have passed the appearance quality inspection, and conduct an in-depth analysis of the accuracy of the inspection work of the inspection post based on the manual spot-check data, thereby improving the timeliness of discovering errors or omissions in the quality inspection process, reducing the risk of product quality problems, and helping enterprises accurately understand the actual effects and existing problems of the quality inspection link.

[0065] The above contents are merely examples and explanations of the concept of the present invention. Those skilled in the art may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the concept of the invention or exceed the scope defined by the present invention, they shall all fall within the protection scope of the present invention.

Claims

1. An AI-vision-based air conditioner appearance intelligent detection data management platform, characterized in that: include: The position correction evaluation module is used to guide each air-conditioning product in the current batch to the inspection position of the target production line through the track in sequence, construct a three-dimensional information model of each air-conditioning product in the current batch at the inspection position, obtain the coordinate value of each feature point on each air-conditioning product in the current batch, and evaluate whether the position of each air-conditioning product in the current batch at the inspection position is corrected. If the position is corrected, the code reader scanning module is executed; if the position is not corrected, the position correction operation is performed; The barcode reader scanning module is used to scan the serial numbers of the air conditioner products in the current batch after the positions of the products are corrected. If the automatic scanning fails, an audible and visual alarm is triggered to guide manual scanning for confirmation. If the automatic scanning succeeds, the air conditioner appearance quality inspection module is executed. The air conditioner appearance quality inspection module is used to collect the quality inspection images corresponding to each inspection surface of each air conditioner product in the current batch, obtain the quality inspection data corresponding to each component in each inspection group in each inspection surface of each air conditioner product from the quality inspection images, and evaluate whether the appearance quality inspection of each air conditioner product in the current batch is qualified. If the appearance quality inspection is qualified, the system displays PASS; if the appearance quality inspection is unqualified, the system displays FAIL and a red light and buzzer alarm sound; A database for storing the reference positions of each characteristic point of the air-conditioning product in the quality inspection equipment, and storing the standard positions of each monitoring point corresponding to each component in each inspection group in each inspection surface of the air-conditioning product; The detection accuracy assessment module is used to record air-conditioning products that have passed the appearance quality inspection as qualified products. Quality inspectors conduct random inspections on qualified products in the current batch, extract the number of unqualified products found in the random inspections, and extract the manual random inspection data corresponding to each historical batch of the target production line to evaluate whether the inspection work of the inspection post in the target production line is accurate. If the inspection work is inaccurate, feedback will be provided immediately.

2. According to claim 1, an AI vision-based air conditioner appearance intelligent detection data management platform is characterized by: The specific process of evaluating whether the position of each air-conditioning product in the current batch in the testing position is correct is as follows: A1. Extract the reference position corresponding to each feature point of the air-conditioning product in the quality inspection equipment from the database, and bring it into the three-dimensional information model to obtain the reference position coordinates corresponding to each feature point of the air-conditioning product in the quality inspection equipment, recorded as ,in, Indicates the number of the feature point. ; A2. Record the coordinate values ​​of each feature point on each air conditioner product in the current batch as ,in, Indicates the product number of the air conditioner. ; A3. Calculate the position offset index of each feature point on each air conditioner product in the current batch , ,in, and Respectively indicate the setting permission axis, Axis and The position offset value corresponding to the axis; A4. Calculate the position deviation index of each air conditioner product in the current batch , ,in, Indicates the number of feature points; A5. Compare the position offset index of each air-conditioning product in the current batch with the position offset index of the set reference. If the position offset index of an air-conditioning product in the current batch is smaller than the position offset index of the set reference, it indicates that the position of the air-conditioning product in the current batch at the inspection post has been corrected. Otherwise, it indicates that the position of the air-conditioning product in the current batch at the inspection post has not been corrected.

3. According to claim 2, an AI vision-based air conditioner appearance intelligent detection data management platform is characterized by: The specific method of performing the position correction operation is: by adjusting the air pressure of the air valve body inside the quality inspection equipment, the push rod is driven to move precisely. When the position of the air-conditioning product moves, the coordinate values ​​of each feature point on the air-conditioning product are collected again, and the position of the air-conditioning product is evaluated again according to the evaluation method of whether the position of the air-conditioning product in the inspection position is corrected. If the position is corrected, the position correction operation is ended. If the position is still not corrected, the position correction operation is continued, and so on, until the position of the air-conditioning product is corrected.

4. According to claim 2, an AI vision-based air conditioner appearance intelligent detection data management platform is characterized by: The quality inspection data includes the location of each monitoring point, the number of external damaged parts and the area of ​​each external damaged part.

5. The AI ​​vision-based air conditioner appearance intelligent detection data management platform according to claim 4 is characterized by: The specific process of evaluating whether the appearance quality inspection of each air-conditioning product in the current batch is qualified is as follows: B1. Extract the location of each monitoring point, the number of appearance damages and the area of ​​each appearance damage from the quality inspection data corresponding to each component in each inspection group on each inspection surface of each air-conditioning product, and calculate the appearance quality inspection pass degree corresponding to each component in each inspection group on each inspection surface of each air-conditioning product ,in, Indicates the number of the detection surface. , Indicates the number of each test group. , Indicates the component number, ; B2. Calculate the appearance quality inspection pass rate corresponding to each test group in each test surface of each air-conditioning product ; B3. Extract the minimum value from the appearance quality inspection pass degree corresponding to each test group in each test surface of each air-conditioning product as the appearance quality inspection pass degree corresponding to each test surface of each air-conditioning product, and compare it with the appearance quality inspection pass degree of the set reference test surface. If the appearance quality inspection pass degree corresponding to any test surface of an air-conditioning product is less than the appearance quality inspection pass degree of the set reference test surface, it indicates that the appearance quality inspection of the air-conditioning product has failed. Otherwise, it indicates that the appearance quality inspection of the air-conditioning product has passed.

6. The AI ​​vision-based air conditioner appearance intelligent detection data management platform according to claim 5 is characterized by: The specific process of calculating the appearance quality inspection pass of each component in each inspection group in each inspection surface of each air-conditioning product is as follows: C1. Based on the positions of each monitoring point, the number of damaged parts and the area of ​​damaged parts corresponding to each component in each test group on each test surface of each air-conditioning product, calculate the placement position qualification of each component in each test group on each test surface of each air-conditioning product. and appearance completeness ; C2. Calculate the appearance quality inspection pass of each component in each test group on each test surface of each air-conditioning product , ,in, and They represent the weights of the set placement position qualification and appearance integrity corresponding to the appearance quality inspection qualification assessment, .

7. The AI ​​vision-based air conditioner appearance intelligent detection data management platform according to claim 5 is characterized by: The specific process of calculating the appearance quality inspection pass corresponding to each inspection group in each inspection surface of each air-conditioning product is as follows: D1. Compare the appearance quality inspection pass of each component in each test group in each test surface of each air-conditioning product with the set reference appearance quality inspection pass. If the appearance quality inspection pass of a component in a certain test group in a certain test surface of an air-conditioning product is less than the set reference appearance quality inspection pass, then the component in the test group in the test surface of the air-conditioning product is recorded as an abnormal component. Count the number of abnormal components in each test group in each test surface of each air-conditioning product and record it as ; D2. Calculate the appearance quality inspection pass rate corresponding to each test group in each test surface of each air-conditioning product , ,in, Indicates the number of components, Indicates the ratio of abnormal components to the set reference.

8. The AI ​​vision-based air conditioner appearance intelligent detection data management platform according to claim 5 is characterized by: The manual inspection data includes the number of qualified air-conditioning products and the number of unqualified air-conditioning products inspected.

9. The AI ​​vision-based air conditioner appearance intelligent detection data management platform according to claim 8, characterized in that: The specific process of evaluating whether the detection work of the detection post in the target production line is accurate is as follows: E1. Count the number of qualified products in the current batch and record it as ; E2. Record the number of unqualified products sampled as ; E3. Calculate the accuracy of the testing work of the testing position in the current batch testing , ,in, represents a natural constant, Indicates the percentage of substandard air-conditioning products used as reference; E4. Extract the number of qualified air-conditioning products and the number of unqualified air-conditioning products from the manual sampling data corresponding to each batch in the history of the target production line. Calculate the accuracy of the inspection work of the inspection post in the inspection of each batch in the history in the same way as the calculation method of the inspection work accuracy of the inspection post in the current batch inspection. ,in, Indicates the number of each historical batch, ; E5. Calculate the accuracy of the inspection work at the inspection post in the target production line ; E6. Compare the detection work accuracy of the detection post in the target production line with the detection work accuracy of the set reference. If the detection work accuracy of the detection post in the target production line is greater than or equal to the detection work accuracy of the set reference, it indicates that the detection work of the detection post in the target production line is accurate. Otherwise, it indicates that the detection work of the detection post in the target production line is inaccurate.

10. The AI ​​vision-based air conditioner appearance intelligent detection data management platform according to claim 9, characterized in that: The specific process of calculating the detection work accuracy of the detection post in the target production line is as follows: F1. Extract the maximum and minimum values ​​from the detection work accuracy of the detection post in each batch of historical detections, and record them as and ; F2. Calculate the accuracy of the inspection work at the inspection post in the target production line , ,in, It means that the testing position is the first in history The accuracy of the testing work in batch testing, and They represent the detection accuracy deviation and the detection accuracy extreme value difference of the set reference respectively. Indicates the number of historical batches.

Images