Optical surface defect detection system based on wall poster coating analysis

Abstract

The invention belongs to the technical field of wall poster coating detection, and particularly relates to an optical surface defect detection system based on wall poster coating analysis, which comprises an optical illumination unit, a wall poster coating imaging unit, a processing identification positioning unit, an operation influence analysis unit, an intelligent control unit and a real-time alarm unit, a detection area is illuminated through the optical illumination unit, the wall poster coating imaging unit collects a wall poster coating surface image in real time, and the processing recognition positioning unit performs defect feature recognition classification based on the wall poster coating surface image and judges whether a defect alarm signal is generated or not, so that the detection efficiency and the detection result accuracy are improved; and the adverse factor influence degree in the wall poster coating defect detection process is analyzed through the operation influence analysis unit and an alarm is given in time, and the process optimization reminding unit carries out process optimization reminding analysis and gives an alarm in time based on defect detection information, so that the wall poster coating production detection supervision difficulty is remarkably reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of wall sticker coating detection, specifically an optical surface defect detection system based on wall sticker coating analysis. Background Art

[0002] Wall sticker coating products are wall decoration materials that use a flexible substrate as a carrier and evenly coat functional materials such as adhesives and decorative coatings on the surface through a coating process, forming wall decoration materials with both decorative and practical functions. During the production process, defects such as black spots, scratches, bubbles, and wrinkles are likely to occur. Therefore, it is necessary to detect surface defects at the end of the production of wall sticker coating products;

[0003] Traditional manual detection methods have problems such as low efficiency, high missed detection rate, and strong subjectivity, making it difficult to meet the requirements of modern industrial production for product quality and production efficiency. Currently, machine vision is mainly used to detect surface defects of wall sticker coatings, but it is still difficult to effectively ensure imaging quality and reduce the difficulty of defect feature extraction, and it is impossible to reasonably judge the urgency of process optimization of the wall sticker coating automated production line and accurately evaluate the production control performance, which is not conducive to reducing the difficulty of production inspection and supervision of wall sticker coatings;

[0004] In view of the above technical defects, a solution is proposed now. Summary of the Invention

[0005] The purpose of the present invention is to provide an optical surface defect detection system based on wall sticker coating analysis, which solves the problems in the prior art that it is difficult to effectively ensure imaging quality and reduce the difficulty of defect feature extraction, and it is impossible to reasonably judge the urgency of process optimization of the wall sticker coating automated production line and accurately evaluate the production control performance, which is not conducive to reducing the difficulty of production inspection and supervision of wall sticker coatings and has a low level of intelligence.

[0006] To achieve the above purpose, the present invention provides the following technical solutions:

[0007] An optical surface defect detection system based on wall sticker coating analysis includes an optical lighting unit, a wall sticker coating imaging unit, a processing and recognition positioning unit, an operation impact analysis unit, an intelligent control unit, and a real-time alarm unit; the intelligent control unit controls the optical lighting unit and the wall sticker coating imaging unit based on the input control data. The optical lighting unit uses a high-brightness LED linear condenser light source to illuminate the detection area. The wall sticker coating imaging unit collects the surface image of the wall sticker coating in real time and sends the surface image of the wall sticker coating to the processing and recognition positioning unit;

[0008] The processing and recognition positioning unit performs denoising, enhancement, grayscale conversion, and binarization on the image of the wall sticker coating surface. Based on deep learning algorithms and traditional image processing algorithms, it extracts the defect features of the wall sticker coating and classifies them. If there are defects exceeding the corresponding allowable requirements, a defect alarm signal is generated and sent to the intelligent control unit. When the intelligent control unit receives the defect alarm signal, the real-time alarm unit issues a corresponding early warning; the operation impact analysis unit analyzes the degree of influence of the adverse factors in the process of wall sticker coating defect detection, and accordingly determines whether to generate an operation impact alarm signal. When the operation impact alarm signal is generated, the real-time alarm unit issues a corresponding early warning through the intelligent control unit.

[0009] Furthermore, the specific analysis process of the operation impact analysis unit includes:

[0010] The conveying speed of the wall sticker coating is collected and marked as the wall sticker output value. The wall sticker output value is numerically compared with the preset wall sticker output value range. If the wall sticker output value is not within the preset wall sticker output value range, it is determined that the wall sticker coating is currently in an abnormal output state; the total duration of the wall sticker coating in the abnormal output state within a unit time is obtained and marked as the wall sticker coating output difference value. Also, the difference between the wall sticker output value and the median of the preset wall sticker output value range is calculated and the absolute value is taken to obtain the output characteristic value. The average value of all output characteristic values within a unit time is calculated to obtain the wall sticker coating output deviation value;

[0011] The wall sticker coating output difference value and the wall sticker coating output deviation value are numerically compared with the preset wall sticker coating output difference threshold and the preset wall sticker coating output deviation threshold respectively. If the wall sticker coating output difference value or the wall sticker coating output deviation value exceeds the corresponding preset threshold, an operation impact alarm signal is generated.

[0012] Furthermore, if both the wall sticker coating output difference value and the wall sticker coating output deviation value do not exceed the corresponding preset thresholds, the deviation value of the real-time brightness of the detection area compared to the set standard brightness value is marked as the area brightness characteristic value, and the deviation degree value of the light color of the detection area compared to the set illumination color is marked as the light color characteristic value. Also, the real-time amplitude of the wall sticker coating jitter is collected and marked as the wall sticker jitter condition value;

[0013] By performing a weighted sum calculation on the area brightness characteristic value, the light color characteristic value, and the wall sticker jitter condition value to obtain the image quality impact value, the image quality impact value is numerically compared with the preset image quality impact threshold. If the image quality impact value exceeds the preset image quality impact threshold, it is determined that the current state is a high-quality impact state;

[0014] Obtain the total duration in the high-quality impact state within a unit time and mark it as the measured value of high-quality impact time. Mark the ratio of the image quality impact value to the preset image quality impact threshold as the impact measurement value, and calculate the average value of all impact measurement values within a unit time to obtain the high-quality impact measurement value. Numerically compare the high-quality impact time measurement value and the high-quality impact measurement value with the preset high-quality impact time measurement threshold and the preset high-quality impact measurement threshold respectively. If the high-quality impact time measurement value or the high-quality impact measurement value exceeds the corresponding preset threshold, generate an operation impact alarm signal.

[0015] Further, the processing and identification positioning unit is communicatively connected to the process optimization reminder unit. The processing and identification positioning unit sends all defect detection information within a unit time to the process optimization reminder unit. If no defect alarm signal is generated within a unit time, the process optimization reminder unit performs process optimization reminder analysis based on the defect detection information to determine whether to generate a process optimization alarm signal, and sends the process optimization alarm signal to the intelligent control unit when the process optimization alarm signal is generated. When the intelligent control unit receives the process optimization alarm signal, the real-time alarm unit issues a corresponding warning.

[0016] Further, the specific analysis process of the process optimization reminder analysis includes:

[0017] Based on all defect detection information within a unit time, obtain the types of all defects that occur. Mark the occurrence times of the corresponding type of defect as the recognition frequency inspection value. Numerically compare the recognition frequency inspection value with the corresponding preset recognition frequency inspection threshold. If the recognition frequency inspection value exceeds the corresponding preset recognition frequency detection threshold, mark the corresponding type of defect as an abnormal type. If there is an abnormal type within a unit time, generate a process optimization alarm signal.

[0018] Further, if there is no abnormal type within a unit time, calculate the ratio of the recognition frequency inspection value of the corresponding type of defect to the corresponding preset recognition frequency inspection threshold to obtain the recognition frequency analysis value. And preset a preset weight value for each type of defect in advance. Multiply the recognition frequency analysis value of the corresponding type of defect by the corresponding preset weight value to obtain the recognition characteristic value.

[0019] And sum up the recognition characteristic values of all types of defects that occur within a unit time to obtain the recognition anomaly value. Numerically compare the recognition anomaly value with the preset recognition anomaly threshold. If the recognition anomaly value exceeds the preset recognition anomaly threshold, generate a process optimization alarm signal.

[0020] Further, the intelligent control unit is communicatively connected to the process control judgment unit. The process control judgment unit is used to set the detection period, analyze the control status of the wall sticker coating production process during the detection period, generate a wall sticker coating production control qualified signal or a wall sticker coating production control abnormal signal through the analysis, and send the wall sticker coating production control qualified signal or the wall sticker coating production control abnormal signal to the intelligent control unit. When the intelligent control unit receives the wall sticker coating production control abnormal signal, it makes the real-time alarm unit issue a corresponding early warning.

[0021] Further, the specific analysis process of the process control judgment unit includes:

[0022] Obtain the generation times of the defect alarm signal and the process optimization alarm signal during the detection period, and mark the sum of the two as the wall sticker coating production anomaly value. Calculate the ratio of the wall sticker coating production anomaly value to the total duration of the wall sticker coating defect detection during the detection period to obtain the quality anomaly alarm value. Compare the quality anomaly alarm value with the preset quality anomaly alarm threshold. If the quality anomaly alarm value exceeds the preset quality anomaly alarm threshold, generate a wall sticker coating production control abnormal signal.

[0023] Further, if the quality anomaly alarm value does not exceed the preset quality anomaly alarm threshold, obtain the production equipment involved in the automated production line of the wall sticker coating, collect the detection data of the operating parameters of the corresponding production equipment, compare the detection data of the operating parameters with the corresponding preset data requirements. When there is detection data of the operating parameters that does not meet the corresponding preset data requirements, it is determined that the corresponding production equipment is in a state to be optimized; start timing when it is determined that the corresponding production equipment is in a state to be optimized until the detection data of the operating parameters of the corresponding production equipment all meet the corresponding preset data requirements, and thus obtain the manual control efficiency value; compare the manual control efficiency value with the corresponding preset manual control efficiency threshold. If the manual control efficiency value exceeds the corresponding preset manual control efficiency threshold, mark the corresponding manual control efficiency value as the manual control inefficiency value;

[0024] Obtain all the manual control efficiency values of the corresponding production equipment during the detection period, sum them up, and calculate the ratio of the sum result to the total production duration of the wall sticker coating automated production line during the detection period to obtain the manual control risk value, and mark the number of the manual control inefficiency values corresponding to the corresponding production equipment during the detection period as the manual control inefficiency count value;

[0025] Obtain the occurrence times of the maintenance interval duration of the corresponding production equipment during the detection period exceeding the corresponding preset interval duration threshold and mark them as the maintenance loss value. Calculate the weighted sum of the manual control risk value, the manual control inefficiency count value, and the maintenance loss value of the corresponding production equipment to obtain the production management matching value. Compare the production management matching value with the corresponding preset production management matching threshold. If the production management matching value exceeds the corresponding preset production management matching threshold, mark the corresponding production equipment as the management non-matching equipment;

[0026] Obtain the ratio of the number of production equipment marked as management non-matching equipment during the detection period and mark it as the non-matching detection value, and mark the ratio of the production management matching value of the corresponding production equipment to the corresponding preset production management matching threshold as the matching detection ratio. Calculate the average value of the matching detection ratios of all production equipment in the wall sticker coating automated production line to obtain the matching measurement value;

[0027] Compare the non-matching detection value and the matching measurement value with the preset non-matching detection threshold and the preset matching measurement threshold respectively. If the non-matching detection value or the matching measurement value exceeds the corresponding preset threshold, generate a wall sticker coating production control abnormal signal; if both the non-matching detection value and the matching measurement value do not exceed the corresponding preset threshold, generate a wall sticker coating production control qualified signal.

[0028] Compared with the prior art, the beneficial effects of the present invention are:

[0029] 1. In the present invention, the defect feature recognition and classification are carried out based on the wall sticker coating surface image by the processing and recognition positioning unit, and it is judged whether to generate a defect alarm signal, so as to ensure the quality of the output wall sticker coating. And the influence degree of the adverse factors in the wall sticker coating defect detection process is analyzed by the operation influence analysis unit to facilitate the smooth and efficient progress of the detection operation. And the process optimization reminder analysis is carried out based on the defect detection information by the process optimization reminder unit to make reasonable optimization and improvement measures in time, ensuring the quality of the produced wall sticker coating and significantly reducing the production supervision difficulty;

[0030] 2. In the present invention, the process control situation of the wall sticker coating production process during the detection period is analyzed by the process control judgment unit to generate a wall sticker coating production control qualified signal or a wall sticker coating production control abnormal signal. When the wall sticker coating production control abnormal signal is generated, the real-time alarm unit issues a corresponding early warning to strengthen the subsequent production management intensity of the wall sticker coating automated production line in time, which is beneficial to reasonably formulate the subsequent production supervision plan, ensuring the high-efficiency production of the wall sticker coating while improving its quality, with a high level of intelligence. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] For the convenience of those skilled in the art to understand, the present invention will be further described below with reference to the accompanying drawings;

[0032] Figure 1 It is the system block diagram of the first embodiment in the present invention;

[0033] Figure 2 It is the system block diagram of the second embodiment in the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0034] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without making creative efforts shall fall within the protection scope of the present invention.

[0035] Embodiment 1: As Figure 1 shown, the optical surface defect detection system based on wall sticker coating analysis proposed by the present invention includes an optical illumination unit, a wall sticker coating imaging unit, a processing and recognition positioning unit, an operation impact analysis unit, a process optimization reminder unit, an intelligent control unit, and a real-time alarm unit;

[0036] The intelligent control unit controls the optical illumination unit and the wall sticker coating imaging unit based on the input control data. The optical illumination unit uses a high-brightness LED linear condenser light source to illuminate the detection area to facilitate the quality of the collected images. It should be noted that for wall sticker coating materials with high light transmittance, a transmissive lighting method (the light source is located below the film and the camera is located above) is combined. For opaque materials, a reflective lighting method (the light source and the camera are on the same side) is used;

[0037] The wall sticker coating imaging unit real-time collects the surface image of the wall sticker coating and sends the surface image of the wall sticker coating to the processing and recognition positioning unit. Among them, the wall sticker coating imaging unit is composed of an industrial CCD camera, a lens, and a photoelectric sensor, and real-time collects the surface image of the wall sticker coating, with a resolution ≥ 4096 × 256 pixels and a frame rate ≥ 800 frames per second.

[0038] The processing and recognition positioning unit performs preprocessing such as denoising, enhancement, grayscale conversion, and binarization on the surface image of the wall sticker coating to improve the defect feature contrast. Based on deep learning algorithms (such as convolutional neural network CNN) and traditional image processing algorithms (such as morphological analysis, edge detection), the defect features (area, length, shape, grayscale value) of the wall sticker coating are extracted and classified into types such as black dots, crystal dots, bubbles, scratches, etc.;

[0039] And after identifying the defects, the various characteristic data of the corresponding defects are compared one by one with the corresponding preset data requirements. If there are defects that exceed the corresponding allowable requirements (that is, the characteristic data of the corresponding defects exceed the corresponding preset data requirements), it indicates that the corresponding defects seriously affect the quality of the wall sticker coating. Then a defect alarm signal is generated and sent to the intelligent control unit. When the intelligent control unit receives the defect alarm signal, it makes the real-time alarm unit issue a corresponding warning to remind the management personnel to pause the detection operation in time and process the wall sticker coating with defects to ensure the quality of the produced wall sticker coating.

[0040] The operation impact analysis unit analyzes the degree of influence of adverse factors in the wall sticker coating defect detection process, and accordingly determines whether to generate an operation impact alarm signal. When generating an operation impact alarm signal, the intelligent control unit enables the real-time alarm unit to issue a corresponding early warning to remind the management personnel to suspend the detection operation and take reasonable improvement measures to ensure the image quality and acquisition efficiency of the surface images of the wall stickers collected, which is conducive to the smooth and efficient progress of the detection operation. The specific analysis process of the operation impact analysis unit is as follows:

[0041] The conveying speed of the wall sticker coating is collected and marked as the wall sticker output value. The wall sticker output value is compared numerically with the preset wall sticker output value range. If the wall sticker output value is not within the preset wall sticker output value range, it indicates that the conveying speed of the wall sticker coating does not meet the requirements, and it is determined that the wall sticker coating is currently in an abnormal output state;

[0042] The total duration of the wall sticker coating in the abnormal output state per unit time is obtained and marked as the wall sticker coating output difference value. Moreover, the difference between the wall sticker output value and the median of the preset wall sticker output value range is calculated and the absolute value is taken to obtain the output characteristic value. The average value of all output characteristic values within the unit time is calculated to obtain the wall sticker coating output deviation value;

[0043] The wall sticker coating output difference value and the wall sticker coating output deviation value are respectively compared numerically with the preset wall sticker coating output difference threshold and the preset wall sticker coating output deviation threshold. If the wall sticker coating output difference value or the wall sticker coating output deviation value exceeds the corresponding preset threshold, it indicates that the conveying performance of the wall sticker coating is poor, which is not conducive to ensuring the detection efficiency, the integrity of the collected images, and the image quality, and an operation impact alarm signal is generated.

[0044] Furthermore, if both the wall sticker coating output difference value and the wall sticker coating output deviation value do not exceed the corresponding preset thresholds, the deviation value of the real-time brightness of the detection area compared with the set standard brightness value is marked as the area brightness characteristic value, and the deviation degree value of the light color of the detection area compared with the set illumination color is marked as the light color characteristic value, and the real-time amplitude of the wall sticker coating jitter is collected and marked as the wall sticker jitter condition value;

[0045] The image quality impact value is obtained by performing a weighted sum calculation on the area brightness characteristic value, the light color characteristic value, and the wall sticker jitter condition value, that is, corresponding preset weight coefficients are assigned to the area brightness characteristic value, the light color characteristic value, and the wall sticker jitter condition value respectively. The area brightness characteristic value, the light color characteristic value, and the wall sticker jitter condition value are respectively multiplied by the corresponding preset weight coefficients, and the sum value of the three product results is marked as the image quality impact value; moreover, the larger the numerical value of the image quality impact value, the more unfavorable it is to ensure the quality of the surface images of the wall stickers collected currently;

[0046] Numerically compare the image quality impact value with a preset image quality impact threshold. If the image quality impact value exceeds the preset image quality impact threshold, it indicates that the current situation is not conducive to ensuring the quality of the collected wall sticker coating surface image, and it is determined that the current is in a high-quality impact state;

[0047] Obtain the total duration in the high-quality impact state within a unit time and mark it as the high-quality impact time measurement value. Mark the ratio of the image quality impact value to the preset image quality impact threshold as the impact occupancy measurement value, and calculate the average value of all impact occupancy measurement values within a unit time to obtain the high-quality impact occupancy measurement value;

[0048] Numerically compare the high-quality impact time measurement value and the high-quality impact occupancy measurement value with the preset high-quality impact time measurement threshold and the preset high-quality impact occupancy measurement threshold respectively. If the high-quality impact time measurement value or the high-quality impact occupancy measurement value exceeds the corresponding preset threshold, it indicates that the detection environment condition of the wall sticker coating is not good and it is difficult to ensure the image quality of the collected image, then an operation impact alarm signal is generated.

[0049] The processing and recognition positioning unit sends all defect detection information within a unit time to the process optimization reminder unit. If no defect alarm signal is generated within a unit time, the process optimization reminder unit performs process optimization reminder analysis based on the defect detection information to determine whether to generate a process optimization alarm signal;

[0050] And when the process optimization alarm signal is generated, it is sent to the intelligent control unit. When the intelligent control unit receives the process optimization alarm signal, it makes the real-time alarm unit issue a corresponding early warning to remind the management personnel to check the wall sticker coating automatic production line in time and make reasonable optimization and improvement measures to ensure the quality of the produced wall sticker coating, significantly reduce the supervision difficulty in the wall sticker coating production process, and has a high level of intelligence; The specific analysis process of the process optimization reminder analysis is as follows:

[0051] Based on all defect detection information within a unit time, obtain the types of all defects that appear. Mark the number of occurrences of the corresponding type of defect as the recognition frequency inspection value. Numerically compare the recognition frequency inspection value with the corresponding preset recognition frequency inspection threshold. If the recognition frequency inspection value exceeds the corresponding preset recognition frequency detection threshold, it indicates that the generation of the corresponding type of defect is relatively frequent, and then mark the corresponding type of defect as an abnormal type; If there is an abnormal type within a unit time, it indicates that the probability of an abnormality in the wall sticker coating automatic production line is relatively large, and then a process optimization alarm signal is generated.

[0052] Furthermore, if there is no abnormal type within a unit time, the recognition frequency analysis value is calculated by taking the ratio of the recognition frequency inspection value of the corresponding type of defect to the corresponding preset recognition frequency inspection threshold. Moreover, a preset weight value is set for each type of defect in advance. The values of the preset weight values are all positive numbers, and the more adverse the impact of the corresponding type of defect on the quality of the wall sticker coating, the larger the value of the preset weight value matched with it. Multiply the recognition frequency analysis value of the corresponding type of defect by the corresponding preset weight value to obtain the recognition characteristic value accordingly.

[0053] Then, sum up the recognition characteristic values of all types of defects occurring within a unit time to obtain the recognition abnormal condition value. Compare the recognition abnormal condition value with the preset recognition abnormal condition threshold. If the recognition abnormal condition value exceeds the preset recognition abnormal condition threshold, it indicates that the probability of an abnormality in the wall sticker coating automated production line is relatively high, and then a process optimization alarm signal is generated.

[0054] Embodiment 2: As Figure 2 shown, the difference between this embodiment and Embodiment 1 is that the intelligent control unit is communicatively connected to the process control judgment unit. The process control judgment unit is used to set the detection period. Preferably, the detection period is fifteen days. Analyze the control status of the wall sticker coating production process within the detection period, and generate a wall sticker coating production control qualified signal or a wall sticker coating production control abnormal signal through the analysis.

[0055] Then, send the wall sticker coating production control qualified signal or the wall sticker coating production control abnormal signal to the intelligent control unit. When the intelligent control unit receives the wall sticker coating production control abnormal signal, it makes the real-time alarm unit issue a corresponding early warning to remind the management personnel to strengthen the subsequent production management intensity of the wall sticker coating automated production line, which is conducive to the management personnel to reasonably formulate the subsequent production supervision plan, ensuring the high-efficiency production of the wall sticker coating while improving its quality. The specific analysis process of the process control judgment unit is as follows:

[0056] Obtain the number of times the defect alarm signal is generated and the number of times the process optimization alarm signal is generated within the detection period, and mark the sum value of the two as the wall sticker coating production abnormality value. Calculate the ratio of the wall sticker coating production abnormality value to the total duration of the wall sticker coating defect detection within the detection period to obtain the quality abnormality alarm value. Compare the quality abnormality alarm value with the preset quality abnormality alarm threshold. If the quality abnormality alarm value exceeds the preset quality abnormality alarm threshold, it indicates that the production quality status of the wall sticker coating within the detection period is poor, and then a wall sticker coating production control abnormal signal is generated.

[0057] Furthermore, if the quality anomaly alarm value does not exceed the preset quality anomaly alarm threshold, obtain the production equipment involved in the automated production line for wall sticker coating, collect the detection data of the operating parameters of the corresponding production equipment, compare the detection data of each operating parameter with the corresponding preset data requirements. When there is detection data of an operating parameter that does not meet the corresponding preset data requirements, it indicates that the corresponding production equipment needs parameter regulation to ensure its safe operation, and then determine that the corresponding production equipment is in a state to be optimized;

[0058] Start timing when it is determined that the corresponding production equipment is in a state to be optimized until the detection data of all operating parameters of the corresponding production equipment meet the corresponding preset data requirements, and thus obtain the manual control efficiency value; compare the manual control efficiency value with the corresponding preset manual control efficiency threshold. If the manual control efficiency value exceeds the corresponding preset manual control efficiency threshold, mark the corresponding manual control efficiency value as a manual control inefficiency value;

[0059] Obtain all the manual control efficiency values of the corresponding production equipment during the detection period and sum them up, and calculate the ratio of the sum value to the total production duration of the wall sticker coating automated production line during the detection period to obtain the manual control risk value, and mark the number of manual control inefficiency values corresponding to the corresponding production equipment during the detection period as the manual control inefficiency count value;

[0060] Obtain the occurrence times of the maintenance interval duration of the corresponding production equipment during the detection period exceeding the corresponding preset interval duration threshold and mark it as the maintenance loss value. Obtain the production management matching value by performing a weighted sum calculation on the manual control risk value, the manual control inefficiency count value, and the maintenance loss value of the corresponding production equipment; that is, assign corresponding preset weight coefficients to the manual control risk value, the manual control inefficiency count value, and the maintenance loss value respectively, multiply the manual control risk value, the manual control inefficiency count value, and the maintenance loss value by the corresponding preset weight coefficients respectively, and mark the sum value of the three product results as the production management matching value; moreover, the larger the value of the production management matching value, the worse the overall management status of the corresponding production equipment;

[0061] Compare the production management matching value with the corresponding preset production management matching threshold. If the production management matching value exceeds the corresponding preset production management matching threshold, indicating that the overall management status of the corresponding production equipment is poor, then mark the corresponding production equipment as a management non - matching equipment;

[0062] Obtain the ratio of the number of production equipment marked as management non - matching equipment during the detection period and mark it as the non - matching detection value, and mark the ratio of the production management matching value of the corresponding production equipment to the corresponding preset production management matching threshold as the matching occupancy detection value. Calculate the average value of the matching occupancy detection values of all production equipment in the wall sticker coating automated production line to obtain the matching occupancy measurement value;

[0063] The non-matching detection value and the matching prediction value are respectively compared numerically with the preset non-matching detection threshold and the preset matching prediction threshold. If the non-matching detection value or the matching prediction value exceeds the corresponding preset threshold, it indicates that the production management status of the wall sticker coating automated production line during the detection period is poor, and a wall sticker coating production control abnormal signal is generated; if both the non-matching detection value and the matching prediction value do not exceed the corresponding preset threshold, it indicates that the overall production management performance of the wall sticker coating automated production line during the detection period is good, and a wall sticker coating production control qualified signal is generated.

[0064] The working principle of the present invention: When in use, the detection area is illuminated by the optical lighting unit, the wall sticker coating imaging unit real-time collects the surface image of the wall sticker coating, the processing and recognition and positioning unit identifies and classifies the defect features based on the surface image of the wall sticker coating and determines whether to generate a defect alarm signal, so as to timely process the wall sticker coating with defects and ensure the quality of the produced wall sticker coating. And the operation influence analysis unit analyzes the influence degree of the adverse factors in the wall sticker coating defect detection process and gives an alarm in time, which is beneficial to the smooth and efficient progress of the detection operation. And the process optimization reminder unit conducts process optimization reminder analysis based on the defect detection information to determine whether to generate a process optimization alarm signal. When the process optimization alarm signal is generated, the wall sticker coating automated production line is inspected and reasonable optimization and improvement measures are taken to ensure the quality of the produced wall sticker coating and significantly reduce the production supervision difficulty, with a high level of intelligence.

[0065] The preferred embodiments of the present invention disclosed above are only used to help illustrate the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the present invention to the specific implementation manners. Obviously, according to the content of this specification, many modifications and changes can be made. These embodiments are selected and specifically described in this specification to better explain the principle and practical application of the present invention, so that those skilled in the art in the relevant technical field can well understand and utilize the present invention. The present invention is only limited by the claims and their full scope and equivalents.

Claims

1. An optical surface defect detection system based on wall sticker coating analysis, characterized in that: It includes an optical lighting unit, a wall sticker coating imaging unit, a processing identification and positioning unit, an operation impact analysis unit, an intelligent control unit and a real-time alarm unit; the intelligent control unit controls the optical lighting unit and the wall sticker coating imaging unit based on the input control data, the optical lighting unit illuminates the detection area, and the wall sticker coating imaging unit collects the wall sticker coating surface image in real time and sends it to the processing identification and positioning unit; The processing, identification and positioning unit processes the surface image of the wall sticker coating, extracts and classifies the defect features of the wall sticker coating based on the deep learning algorithm and the traditional image processing algorithm. If there are defects that exceed the corresponding allowable requirements, a defect alarm signal is generated and sent to the intelligent control unit; the operation impact analysis unit analyzes the degree of influence of adverse factors in the wall sticker coating defect detection process, and when the operation impact alarm signal is generated, the real-time alarm unit issues a corresponding warning through the intelligent control unit.

2. The optical surface defect detection system based on wall sticker coating analysis according to claim 1 is characterized in that: The specific analysis process of the operation impact analysis unit includes: The total duration of the wall sticker coating in the output abnormal state per unit time is obtained and marked as the wall sticker coating input anomaly value, and the mean of all output characteristic values ​​per unit time is calculated to obtain the wall sticker coating input deviation value; if the wall sticker coating input anomaly value or the wall sticker coating input deviation value exceeds the corresponding preset threshold, an operation impact alarm signal is generated.

3. The optical surface defect detection system based on wall sticker coating analysis according to claim 2 is characterized in that: If the wall sticker coating input difference value and the wall sticker coating input deviation value do not exceed the corresponding preset threshold value, the image quality impact value is calculated by weighted summing the area brightness feature value, the light color feature value and the wall sticker jitter value. If the image quality impact value exceeds the preset image quality impact threshold, it is judged that the current state is in high quality impact; The total time in the high quality impact state per unit time is obtained and marked as the high quality impact time measurement value, and the ratio of the image quality impact value to the preset image quality impact threshold is marked as the impact share measurement value, and the average of all the impact share measurement values ​​per unit time is calculated to obtain the high quality impact share measurement value; if the high quality impact time measurement value or the high quality impact share measurement value exceeds the corresponding preset threshold, an operation impact alarm signal is generated.

4. The optical surface defect detection system based on wall sticker coating analysis according to claim 1 is characterized in that: The processing identification and positioning unit is communicatively connected to the process optimization reminder unit. The processing identification and positioning unit sends all defect detection information within a unit time to the process optimization reminder unit. If no defect alarm signal is generated within the unit time, the process optimization reminder unit performs a process optimization reminder analysis based on the defect detection information, and sends the process optimization alarm signal to the intelligent control unit when it is generated.

5. The optical surface defect detection system based on wall sticker coating analysis according to claim 4 is characterized in that: The specific analysis process of process optimization reminder analysis is as follows: based on all defect detection information within a unit time, the types of all defects that occur are obtained, and the number of occurrences of defects of the corresponding type is marked as the identification frequency detection value. If the identification frequency detection value exceeds the corresponding preset identification frequency detection threshold, the corresponding type of defect is marked as an abnormal type; if an abnormal type exists, a process optimization alarm signal is generated.

6. The optical surface defect detection system based on wall sticker coating analysis according to claim 5 is characterized in that: If there is no abnormal type within a unit time, the identification frequency detection value of the corresponding type of defect is calculated by ratio with the corresponding preset identification frequency detection threshold to obtain the identification frequency analysis value, and the identification frequency analysis value of the corresponding type of defect is multiplied by the corresponding preset weight value to obtain the identification characteristic value; and the identification characteristic values ​​of all types of defects appearing within a unit time are summed up to obtain the identification abnormality value. If the identification abnormality value exceeds the preset identification abnormality threshold, a process optimization alarm signal is generated.

7. The optical surface defect detection system based on wall sticker coating analysis according to claim 1 is characterized in that: The intelligent control unit is communicatively connected to the process control and judgment unit, which is used to set a detection period, analyze the control status of the wall sticker coating production process during the detection period, generate a wall sticker coating production control qualified signal or a wall sticker coating production control abnormal signal through analysis, and send the wall sticker coating production control qualified signal or the wall sticker coating production control abnormal signal to the intelligent control unit.

8. The optical surface defect detection system based on wall sticker coating analysis according to claim 7 is characterized in that: The specific analysis process of the process control judgment unit includes: The number of defect alarm signals and process optimization alarm signals generated during the detection period are obtained and the sum of the two is marked as the wall sticker coating production abnormality value. The wall sticker coating production abnormality value is ratioed with the total duration of wall sticker coating defect detection during the detection period to obtain the quality abnormality alarm value. If the quality abnormality alarm value exceeds the preset quality abnormality alarm threshold, a wall sticker coating production control abnormality signal is generated.

9. The optical surface defect detection system based on wall sticker coating analysis according to claim 8 is characterized in that: If the quality deviation alarm value does not exceed the preset quality deviation alarm threshold, the proportion of production equipment marked as managed non-matching equipment during the detection period is obtained and marked as a non-matching detection value, and the matching proportion detection values ​​of all production equipment in the wall sticker coating automated production line are averaged to obtain the matching proportion measurement value; if the non-matching detection value or the matching proportion measurement value exceeds the corresponding preset threshold, a wall sticker coating production control abnormal signal is generated; otherwise, a wall sticker coating production control qualified signal is generated.

Images