Liquid crystal panel AOI detection system based on artificial intelligence

Abstract

The invention discloses a liquid crystal panel AOI detection system based on artificial intelligence, relates to the technical field of liquid crystal panel detection, solves the technical problem that in the prior art, a liquid crystal panel cannot be pointedly overhauled according to different defect type analysis, and particularly relates to defect type analysis on a liquid crystal panel detection result, so that the detection accuracy of the liquid crystal panel is improved. The method comprises the following steps: acquiring a detection surface of a liquid crystal panel according to a real-time image, comparing actual brightness of each position in the detection surface with a set brightness threshold range, marking positions not in the set brightness threshold range as abnormal point locations, and combining and constructing an abnormal surface according to the positions of the abnormal point locations. If there is no adjacent abnormal point location, marking the corresponding abnormal point location as an abnormal point; the defect types of the liquid crystal panel are analyzed through liquid crystal panel detection, and targeted defect maintenance is carried out on the liquid crystal panel according to different types, so that the detection efficiency and the defect maintenance timeliness of the liquid crystal panel are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of liquid crystal panel detection, and in particular to an artificial intelligence-based liquid crystal panel AOI detection system. Background Art

[0002] LCD panel AOI inspection is a technology used for quality inspection during the production process of LCD panels. At present, LCD panel AOI inspection technology has been widely used in the LCD panel manufacturing industry; almost all large-scale LCD panel manufacturers have adopted AOI inspection systems to ensure product quality; with the continuous expansion of the LCD panel market and the continuous advancement of technology, the market demand for AOI inspection systems is also continuing to grow.

[0003] However, in the prior art, it is not possible to perform performance testing on the auxiliary detection device during LCD panel testing, and the test quality cannot be guaranteed. It is also not possible to perform device movement testing and operation analysis after determining that the auxiliary detection device is qualified, resulting in a decrease in the accuracy of LCD panel testing. In addition, it is not possible to perform targeted inspection of the LCD panel based on analysis of different defect types.

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

[0005] The purpose of the present invention is to solve the above-mentioned problems and to propose an artificial intelligence-based liquid crystal panel AOI inspection system.

[0006] The purpose of the present invention can be achieved through the following technical solutions:

[0007] An artificial intelligence-based liquid crystal panel AOI detection system includes a liquid crystal panel detection platform, the liquid crystal panel detection platform is communicatively connected to a detection device evaluation unit, wherein the detection device evaluation unit is connected to:

[0008] The detection auxiliary device detection unit is used to perform performance detection on the liquid crystal panel detection auxiliary device, collect detection information of the detection auxiliary device, substitute it into the formula to obtain the detection coefficient of the detection auxiliary device in the liquid crystal panel detection process, and infer whether the detection of the detection auxiliary device is normal according to the comparison of the coefficient threshold;

[0009] A detection movement execution analysis unit is used to perform execution analysis on the detection movement of the detection auxiliary device, obtain trajectory deviation information and movement clarity information, and infer whether the detection movement execution analysis is qualified based on information comparison;

[0010] The defect type analysis unit is used to analyze the defect type of the liquid crystal panel detection result, collect the detection surface of the liquid crystal panel according to the real-time image, compare the actual brightness of each position in the detection surface with the set brightness threshold range, mark the position that is not within the set brightness threshold range as an abnormal point, and merge and construct an abnormal surface according to the position of the abnormal point. If there is no adjacent abnormal point, the corresponding abnormal point will be marked as an abnormal point;

[0011] The test risk assessment unit is used to perform test risk assessment on the LCD panel detection process, collect vibration impact data and light impact data, and infer whether the test risk assessment is normal based on data analysis.

[0012] As a preferred embodiment of the present invention, the detection information of the detection auxiliary device includes the maximum range of the area adapted to the different specifications of LCD panels during the detection process of all current specifications of LCD panels and the corresponding range area deviation value of the maximum range of the lens focus adjustment field angle during the detection process, the maximum deviation value of the image resolution corresponding to adjacent moments during the camera lens focal length change when the electronic component area array arrangement and the linear array arrangement of the different specifications of LCD panels in the detection process of all current specifications of LCD panels alternate, and the deviation of the light intensity uniformity value corresponding to the light intensity uniformity value of each position in the detection area corresponding to the fixed angle of the real-time light source in the detection area and the light intensity uniformity value of each position in the detection area corresponding to the moving angle when different specifications of LCD panels are performing camera image acquisition.

[0013] As a preferred embodiment of the present invention, if the detection coefficient of the detection auxiliary device during the liquid crystal panel detection process exceeds the detection coefficient threshold, it is determined that the detection of the detection auxiliary device during the liquid crystal panel detection process is abnormal; if the detection coefficient of the detection auxiliary device during the liquid crystal panel detection process does not exceed the detection coefficient threshold, it is determined that the detection of the detection auxiliary device during the liquid crystal panel detection process is normal.

[0014] As a preferred embodiment of the present invention, the trajectory deviation information and the moving clarity information are respectively the distance deviation value between the actual moving trajectory of the LCD panel to be detected corresponding to the beginning and end moments of the moving phase of the LCD panel to be detected during the LCD panel detection process and the camera shooting trajectory, and the clarity average of the image subframe photos corresponding to the shooting moments at each moving moment point in the moving phase of the LCD panel to be detected during the LCD panel detection process and the excess of the required clarity for detection.

[0015] As a preferred embodiment of the present invention, if the trajectory deviation information exceeds the trajectory distance deviation value threshold, or the movement clarity information exceeds the clarity excess threshold, then it is determined that the detection movement execution analysis during the LCD panel detection process is unqualified; if the trajectory deviation information does not exceed the trajectory distance deviation value threshold, and the movement clarity information does not exceed the clarity excess threshold, then it is determined that the detection movement execution analysis during the LCD panel detection process is qualified.

[0016] As a preferred embodiment of the present invention, the abnormal points are analyzed:

[0017] The excess value between the brightness of the abnormal point and the peak value of the set brightness threshold range is obtained. If the corresponding excess value is a positive number, the abnormal point is marked as a highlighted abnormal point. At the same time, a circuit maintenance signal is generated and sent together with the circuit maintenance signal and the highlighted abnormal point to the LCD panel detection platform. The LCD panel detection platform performs maintenance on the highlighted abnormal point.

[0018] On the contrary, if the corresponding excess value is a positive number, the abnormal point is marked as a low-brightness abnormal point, a panel inspection signal is generated, and the panel inspection signal is sent to the liquid crystal panel detection platform.

[0019] As a preferred embodiment of the present invention, the abnormal surface is analyzed:

[0020] The brightness deviation values ​​of adjacent abnormal points in the abnormal surface are obtained. If the corresponding brightness deviation value exceeds the brightness deviation threshold, and each adjacent abnormal point constructs a nonlinear abnormal surface, a material mismatch signal is generated; the LCD panel detection platform sends a maintenance instruction and arranges the LCD panel design administrator to conduct an inspection;

[0021] If the corresponding brightness deviation value exceeds the brightness deviation threshold, and each adjacent abnormal point constructs a linear abnormal surface, a traceability management signal is generated, and the LCD panel inspection platform sends a traceability instruction to the production line and arranges the production line supervisor to conduct a traceability inspection.

[0022] As a preferred embodiment of the present invention, the vibration influence data and the light influence data are respectively the vibration frequency generated by the surrounding environment on the detection auxiliary device when the detection area image is collected during the liquid crystal panel detection process, and the floating amount of the numerical deviation of the real-time ambient light intensity corresponding to the surrounding environment and the fill light source of the detection auxiliary device when the detection area image is collected during the liquid crystal panel detection process.

[0023] As a preferred embodiment of the present invention, if the vibration impact data exceeds the vibration frequency generation amount threshold, or the light impact data exceeds the light intensity value deviation floating amount threshold, it is determined that the test risk assessment during the liquid crystal panel detection process is abnormal;

[0024] If the vibration impact data does not exceed the vibration frequency generation threshold, and the light impact data does not exceed the light intensity value deviation floating threshold, it is determined that the test risk assessment during the LCD panel inspection is normal.

[0025] Compared with the prior art, the present invention has the following beneficial effects:

[0026] 1. In the present invention, the performance of the liquid crystal panel detection auxiliary device is tested, and whether the current liquid crystal panel AOI test is qualified is inferred by evaluating the performance of the detection auxiliary device, so as to avoid the abnormal performance of the detection auxiliary device during the liquid crystal panel detection process, which leads to deviation in the liquid crystal panel detection result; the detection efficiency of the liquid crystal panel can be improved by detecting the performance of the detection auxiliary device, and at the same time, when the detection auxiliary device is abnormal, it can be adjusted and replaced in time without affecting the detection progress of the liquid crystal panel;

[0027] During the LCD panel inspection process, the detection movement of the detection auxiliary device is analyzed to infer whether the detection movement execution of the detection auxiliary device meets the actual needs, to ensure that there is no image omission in the device detection movement execution, and to avoid the LCD panel image acquisition time being too long when the device detection movement is in progress, resulting in unnecessary cost waste; through the detection movement execution analysis, the efficiency of LCD panel inspection can be guaranteed while improving the resource utilization of the detection equipment.

[0028] 2. In the present invention, the defect type analysis is performed on the liquid crystal panel detection results, and the defect type of the liquid crystal panel is analyzed through liquid crystal panel detection, and targeted defect maintenance of the liquid crystal panel is performed according to different types, so as to improve the detection efficiency of the liquid crystal panel and the timeliness of defect repair;

[0029] Conduct test risk assessment on the LCD panel inspection process, and infer whether there are risks in the current LCD panel inspection through analysis of the LCD panel inspection process, so that timely maintenance and control can be carried out after any impact occurs during the LCD panel inspection process, thereby reducing the risk of abnormal LCD panel equipment and avoiding abnormal execution of the LCD panel inspection process, which leads to unnecessary costs in the inspection process. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to facilitate understanding by those skilled in the art, the present invention is further described below with reference to the accompanying drawings.

[0031] Figure 1 It is the overall principle block diagram of the present invention;

[0032] Figure 2 This is a principle block diagram of Embodiment 1 of the present invention;

[0033] Figure 3 This is a principle block diagram of embodiment 3 of the present invention. DETAILED DESCRIPTION

[0034] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings 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.

[0035] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0036] See also Figure 1 As shown, an artificial intelligence-based liquid crystal panel AOI inspection system includes a liquid crystal panel inspection platform, wherein the liquid crystal panel inspection platform is communicatively connected to a detection equipment evaluation unit and a defect type analysis unit;

[0037] Embodiment 1

[0038] This embodiment performs equipment testing on the liquid crystal panel testing process to ensure the operation qualification of the testing equipment. Figure 2 As shown, the liquid crystal panel inspection platform is communicatively connected with the inspection equipment evaluation unit, and is provided with an inspection auxiliary device inspection unit and an inspection movement execution analysis unit;

[0039] When the liquid crystal panel detection is executed, after the liquid crystal panel detection platform generates a detection equipment evaluation instruction to the detection equipment evaluation unit, the detection equipment evaluation unit generates a detection auxiliary device detection signal and sends the detection auxiliary device detection signal to the detection auxiliary device detection unit. After the detection auxiliary device detection unit receives the detection auxiliary device detection signal, it performs a performance test on the liquid crystal panel detection auxiliary device, and infers whether the current liquid crystal panel AOI detection is qualified through the performance evaluation of the detection auxiliary device, thereby avoiding abnormal performance of the detection auxiliary device during the liquid crystal panel detection process, which leads to deviations in the liquid crystal panel detection results; the detection efficiency of the liquid crystal panel can be improved by the performance test of the detection auxiliary device, and at the same time, when the detection auxiliary device is abnormal, it can be adjusted and replaced in time without affecting the detection progress of the liquid crystal panel;

[0040] Obtain the maximum range of the LCD panels of different specifications that are adapted to the detection process of all current specifications of the LCD panels and the corresponding range area deviation value of the maximum range of the field of view angle of the lens focusing adjustment during the detection process, and at the same time, obtain the maximum deviation value of the image resolution corresponding to adjacent moments during the focal length transformation of the camera lens when the electronic components corresponding to the LCD panels of different specifications in the detection process of all current specifications of the LCD panels are arranged alternately with the surface array arrangement and the linear array arrangement, and set the maximum range of the LCD panels of different specifications that are adapted to the detection process of all current specifications of the LCD panels and the corresponding range area deviation value of the maximum range of the field of view angle of the lens focusing adjustment during the detection process, and the maximum deviation value of the image resolution corresponding to adjacent moments during the focal length transformation of the camera lens when the electronic components corresponding to the surface array arrangement and the linear array arrangement are arranged alternately during the detection process of all current specifications of the LCD panels, and set the labels QYP and FBP respectively;

[0041] When different specifications of LCD panels are obtained for camera image acquisition, the light intensity average value of each position in the detection area corresponding to the fixed angle of the real-time light source in the detection area and the light intensity average value of each position in the detection area corresponding to the moving angle correspond to the light intensity average value deviation, and when different specifications of LCD panels are used for camera image acquisition, the light intensity average value of each position in the detection area corresponding to the fixed angle of the real-time light source in the detection area and the light intensity average value of each position in the detection area corresponding to the moving angle correspond to the light intensity average value deviation as SZB; wherein the light intensity average value is represented by the average value of the actual light intensity corresponding to each position in the detection area;

[0042] The above collected data are uniformly marked as the detection information of the detection auxiliary device, and substituted into the formula to obtain the detection coefficient of the detection auxiliary device in the liquid crystal panel detection process, where the formula is:

[0043] Wherein, JC is the detection coefficient of the detection auxiliary device, cmj1, cmj2 and cmj3 are the preset proportional coefficients of the range area deviation, the resolution deviation peak and the corresponding light intensity uniformity deviation, respectively, and e is a natural constant, and α is an error correction factor with a value of 0.97;

[0044] Compare the detection coefficient of the detection auxiliary device with the detection coefficient threshold during the LCD panel detection process:

[0045] If the detection coefficient of the detection auxiliary device during the liquid crystal panel detection process exceeds the detection coefficient threshold, it is determined that the detection auxiliary device during the liquid crystal panel detection process is abnormal, and a device performance deviation signal is generated and sent to the detection equipment platform and forwarded to the liquid crystal panel detection platform. After the liquid crystal panel detection platform receives the device performance deviation signal, the performance of the detection auxiliary device during the liquid crystal panel detection process is debugged or the device is replaced, wherein the detection auxiliary device in the present application includes devices such as cameras and lenses;

[0046] If the detection coefficient of the detection auxiliary device during the liquid crystal panel detection process does not exceed the detection coefficient threshold, it is determined that the detection of the detection auxiliary device during the liquid crystal panel detection process is normal, and a device performance compliance signal is generated and sent to the detection equipment platform and forwarded to the liquid crystal panel detection platform;

[0047] At the same time, a detection movement execution analysis signal is generated and sent to the detection movement execution analysis unit. After receiving the detection movement execution analysis signal, the detection movement execution analysis unit performs execution analysis on the detection movement of the detection auxiliary device during the liquid crystal panel detection process, and infers whether the detection movement execution of the detection auxiliary device meets the actual demand, so as to ensure that there is no image leakage in the detection movement execution of the device, and also avoid that the image acquisition time of the liquid crystal panel is too long when the device detects the movement, resulting in unnecessary cost waste; through the detection movement execution analysis, the detection efficiency of the liquid crystal panel can be guaranteed while the resource utilization of the detection equipment can be improved;

[0048] The distance deviation value between the actual moving trajectory of the LCD panel to be detected corresponding to the beginning and end moments of the moving stage of the LCD panel to be detected and the camera shooting trajectory is obtained during the LCD panel detection process. At the same time, the clarity average of the image subframe photos corresponding to the shooting moments at each moving time point in the moving stage of the LCD panel to be detected during the LCD panel detection process and the excess of the detection required clarity are obtained. The distance deviation value between the actual moving trajectory of the LCD panel to be detected corresponding to the beginning and end moments of the moving stage of the LCD panel to be detected during the LCD panel detection process and the camera shooting trajectory, the clarity average of the image subframe photos corresponding to the shooting moments at each moving time point in the moving stage of the LCD panel to be detected during the LCD panel detection process and the excess of the detection required clarity are marked as trajectory deviation information and movement clarity information, respectively, and compared with the trajectory distance deviation value threshold and the clarity excess threshold, respectively:

[0049] If the distance deviation between the actual moving trajectory of the LCD panel to be detected corresponding to the beginning and end moments of the moving phase of the LCD panel to be detected and the camera shooting trajectory during the LCD panel detection process exceeds the trajectory distance deviation value threshold, or the excess of the clarity of the image subframe photos corresponding to the shooting moments at each moving moment point in the moving phase of the LCD panel to be detected during the LCD panel detection process and the clarity required for detection exceeds the clarity excess threshold, then it is determined that the detection movement execution analysis during the LCD panel detection process is unqualified, an execution movement abnormality signal is generated and sent to the detection equipment platform and forwarded to the LCD panel detection platform. After receiving the signal, the detection auxiliary device corresponding to the LCD panel detection is controlled to move to ensure that the device moving node is consistent with the LCD panel moving node, and the image acquisition of the corresponding node meets the actual acquisition requirements;

[0050] If the distance deviation between the actual moving trajectory of the LCD panel to be detected corresponding to the beginning and end moments of the moving phase of the LCD panel to be detected and the camera shooting trajectory during the LCD panel detection process does not exceed the trajectory distance deviation value threshold, and the excess of the clarity of the image subframe photos corresponding to the shooting moments at each moving moment point in the moving phase of the LCD panel to be detected during the LCD panel detection process and the required clarity does not exceed the excess clarity threshold, then it is determined that the detection movement execution analysis during the LCD panel detection process is qualified, and a normal execution movement signal is generated and sent to the detection equipment platform and forwarded to the LCD panel detection platform;

[0051] Embodiment 2

[0052] See also Figure 1-2 As shown, the LCD panel inspection platform is communicatively connected to a defect type analysis unit. After completing the inspection equipment evaluation, a defect type analysis signal is generated and sent to the defect type analysis unit. After receiving the defect type analysis signal, the defect type analysis unit performs defect type analysis on the LCD panel inspection result. The defect type of the LCD panel is analyzed through LCD panel inspection, and targeted defect maintenance of the LCD panel is performed according to different types, so as to improve the inspection efficiency of the LCD panel and the timeliness of defect repair.

[0053] According to the detection surface of the LCD panel collected by the real-time image, the actual brightness of each position in the detection surface is compared with the set brightness threshold range, and the position that is not within the set brightness threshold range is marked as an abnormal point, and the abnormal point position is merged to construct an abnormal surface according to the position of the abnormal point position. If there is no adjacent abnormal point position, the corresponding abnormal point position is marked as an abnormal point;

[0054] The abnormal point is analyzed to obtain the excess value between the brightness of the abnormal point and the peak value of the set brightness threshold range. If the corresponding excess value is a positive number, the abnormal point is marked as a highlighted abnormal point. At the same time, a circuit maintenance signal is generated and the circuit maintenance signal and the highlighted abnormal point are sent to the LCD panel detection platform. The LCD panel detection platform performs maintenance on the highlighted abnormal point, that is, performs circuit maintenance on the highlighted abnormal point. It can be understood that the main cause of the highlighted point is circuit failure or poor electrode contact. The LCD panel detection platform sends a maintenance instruction and arranges an electrical administrator to perform maintenance.

[0055] On the contrary, if the corresponding excess value is a positive number, the abnormal point is marked as a low-brightness abnormal point, a panel inspection signal is generated and the panel inspection signal is sent to the LCD panel inspection platform. After the LCD panel inspection platform receives the panel inspection signal, it sends an inspection instruction and arranges the panel assembly manager to conduct an inspection;

[0056] The abnormal surface is analyzed to obtain the brightness deviation value of the adjacent abnormal points in the abnormal surface. If the corresponding brightness deviation value exceeds the brightness deviation threshold, and each adjacent abnormal point constructs a nonlinear abnormal surface, a material mismatch signal is generated and sent to the liquid crystal panel detection platform. After receiving the material mismatch signal, the liquid crystal panel detection platform sends a maintenance instruction and arranges the liquid crystal panel design administrator to check to infer whether the liquid crystal materials of the liquid crystal panel are evenly distributed or whether the backlight source is evenly irradiated;

[0057] If the corresponding brightness deviation value exceeds the brightness deviation threshold, and each adjacent abnormal point constructs a linear abnormal surface, a traceability management signal is generated and sent to the LCD panel detection platform. After receiving the traceability management signal, the LCD panel detection platform sends a traceability instruction to the production line and arranges the production line supervisor to conduct a traceability inspection. According to the full process of the LCD panel, the traceability is monitored and traced. While repairing the scratches on the current LCD panel, the location where the scratches are generated is traced to facilitate timely avoidance and reduce the efficiency of scratch generation.

[0058] Embodiment 3

[0059] See also Figure 3 As shown, the LCD panel inspection platform is communicatively connected with a test risk assessment unit. After completing the defect type analysis, a test risk assessment signal is generated and sent to the test risk assessment unit. After receiving the test risk assessment signal, the test risk assessment unit performs a test risk assessment on the LCD panel inspection process. Through the analysis of the LCD panel inspection process, it is inferred whether there is a risk in the current LCD panel inspection, so that maintenance and control can be carried out in time after an impact occurs in the LCD panel inspection process, thereby reducing the risk of abnormal LCD panel equipment and avoiding abnormal execution of the LCD panel inspection process, which leads to unnecessary costs in the inspection process.

[0060] The vibration frequency generated by the surrounding environment on the auxiliary detection device when the detection area image is collected during the LCD panel detection process is obtained, and at the same time, the floating amount of the numerical deviation of the surrounding environment and the fill light source of the detection auxiliary device corresponding to the real-time ambient light intensity when the detection area image is collected during the LCD panel detection process is obtained, and the vibration frequency generated by the surrounding environment on the auxiliary detection device when the detection area image is collected during the LCD panel detection process and the floating amount of the numerical deviation of the surrounding environment and the fill light source of the detection auxiliary device corresponding to the real-time ambient light intensity when the detection area image is collected during the LCD panel detection process are marked as vibration impact data and light impact data, respectively, and compared with the vibration frequency generation threshold and the light intensity value deviation floating threshold, respectively:

[0061] If the vibration frequency generated by the surrounding environment on the detection auxiliary device when the detection area image is collected during the LCD panel detection process exceeds the vibration frequency generation threshold, or the floating amount of the numerical deviation of the real-time ambient light intensity corresponding to the fill light source of the detection auxiliary device when the detection area image is collected during the LCD panel detection process exceeds the light intensity numerical deviation floating amount threshold, then it is determined that the test risk assessment during the LCD panel detection process is abnormal, and a test risk assessment abnormal signal is generated and sent to the LCD panel detection platform. After the LCD panel detection platform receives the test risk assessment abnormal signal, it re-detects the LCD panel being detected at the current moment, and controls the surrounding environment of the detection area to reduce the impact of the surrounding environment on the detection area;

[0062] If the vibration frequency generated by the surrounding environment on the detection auxiliary device when the detection area image is collected during the LCD panel detection process does not exceed the vibration frequency generation threshold, and the floating amount of the numerical deviation of the real-time ambient light intensity corresponding to the surrounding environment and the fill light source of the detection auxiliary device when the detection area image is collected during the LCD panel detection process does not exceed the light intensity numerical deviation floating amount threshold, then it is determined that the test risk assessment during the LCD panel detection process is normal, and a test risk assessment normal signal is generated and the test risk assessment normal signal is sent to the LCD panel detection platform.

[0063] The above formulas are obtained by collecting a large amount of data and performing software simulation, and a formula close to the actual value is selected. The coefficients in the formula are set by technicians in this field according to actual conditions;

[0064] When the present invention is in use, the detection auxiliary device detection unit performs performance detection on the liquid crystal panel detection auxiliary device, collects detection auxiliary device detection information, obtains the detection auxiliary device detection coefficient during the liquid crystal panel detection process, and infers whether the detection of the detection auxiliary device is normal based on the coefficient threshold comparison; the detection movement execution analysis unit performs execution analysis on the detection movement of the detection auxiliary device, obtains trajectory deviation information and movement clarity information, and infers whether the detection movement execution analysis is qualified based on the information comparison; the defect type analysis unit performs defect type analysis on the liquid crystal panel detection result; the test risk assessment unit performs test risk assessment on the liquid crystal panel detection process, collects vibration impact data and light impact data, and infers whether the test risk assessment is normal based on the data analysis.

[0065] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only specific implementation methods. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. An artificial intelligence-based liquid crystal panel AOI inspection system, characterized in that: It includes a liquid crystal panel detection platform, the liquid crystal panel detection platform is communicatively connected with a detection equipment evaluation unit, wherein the detection equipment evaluation unit is connected with: The detection auxiliary device detection unit is used to perform performance detection on the liquid crystal panel detection auxiliary device, collect detection information of the detection auxiliary device, substitute it into the formula to obtain the detection coefficient of the detection auxiliary device in the liquid crystal panel detection process, and infer whether the detection of the detection auxiliary device is normal according to the comparison of the coefficient threshold; A detection movement execution analysis unit is used to perform execution analysis on the detection movement of the detection auxiliary device, obtain trajectory deviation information and movement clarity information, and infer whether the detection movement execution analysis is qualified based on information comparison; The defect type analysis unit is used to analyze the defect type of the liquid crystal panel detection result, collect the detection surface of the liquid crystal panel according to the real-time image, compare the actual brightness of each position in the detection surface with the set brightness threshold range, mark the position that is not within the set brightness threshold range as an abnormal point, and merge and construct an abnormal surface according to the position of the abnormal point. If there is no adjacent abnormal point, the corresponding abnormal point will be marked as an abnormal point; The test risk assessment unit is used to perform test risk assessment on the LCD panel detection process, collect vibration impact data and light impact data, and infer whether the test risk assessment is normal based on data analysis.

2. The liquid crystal panel AOI inspection system based on artificial intelligence according to claim 1, characterized in that: The detection information of the detection auxiliary device includes the maximum range of the area adapted to the LCD panels of different specifications in the detection process of all current specifications of LCD panels and the corresponding range area deviation value of the maximum range of the lens focus adjustment field angle in the detection process, the maximum deviation value of the image resolution corresponding to adjacent moments in the process of camera lens focal length change when the electronic component area array arrangement and line array arrangement of the LCD panels of different specifications in the detection process of all current specifications of LCD panels alternate, and the deviation of the light intensity uniformity value at each position in the detection area corresponding to the fixed angle of the real-time light source in the detection area and the light intensity uniformity value at each position in the detection area corresponding to the moving angle when different specifications of LCD panels are used for camera image acquisition.

3. The artificial intelligence-based liquid crystal panel AOI inspection system according to claim 2, characterized in that: If the detection coefficient of the detection auxiliary device during the liquid crystal panel detection process exceeds the detection coefficient threshold, it is determined that the detection auxiliary device during the liquid crystal panel detection process is abnormal; If the detection coefficient of the detection auxiliary device during the liquid crystal panel detection process does not exceed the detection coefficient threshold, it is determined that the detection of the detection auxiliary device during the liquid crystal panel detection process is normal.

4. The artificial intelligence-based liquid crystal panel AOI inspection system according to claim 1, characterized in that: The trajectory deviation information and the moving clarity information are respectively the distance deviation value between the actual moving trajectory of the LCD panel to be detected corresponding to the beginning and end moments of the moving stage of the LCD panel to be detected during the LCD panel detection process and the camera shooting trajectory, and the average clarity of the image sub-frame photos corresponding to the shooting moments at each moving moment point in the moving stage of the LCD panel to be detected during the LCD panel detection process and the excess of the required clarity for detection.

5. The artificial intelligence-based liquid crystal panel AOI inspection system according to claim 4, characterized in that: If the trajectory deviation information exceeds the trajectory distance deviation value threshold, or the motion clarity information exceeds the clarity excess threshold, it is determined that the detection movement execution analysis during the LCD panel detection process is unqualified; if the trajectory deviation information does not exceed the trajectory distance deviation value threshold, and the motion clarity information does not exceed the clarity excess threshold, it is determined that the detection movement execution analysis during the LCD panel detection process is qualified.

6. The artificial intelligence-based liquid crystal panel AOI inspection system according to claim 1, characterized in that: Analyze the outliers: The excess value between the brightness of the abnormal point and the peak value of the set brightness threshold range is obtained. If the corresponding excess value is a positive number, the abnormal point is marked as a highlighted abnormal point. At the same time, a circuit maintenance signal is generated and sent together with the circuit maintenance signal and the highlighted abnormal point to the LCD panel detection platform. The LCD panel detection platform performs maintenance on the highlighted abnormal point. On the contrary, if the corresponding excess value is a positive number, the abnormal point is marked as a low-brightness abnormal point, a panel inspection signal is generated, and the panel inspection signal is sent to the liquid crystal panel detection platform.

7. The artificial intelligence-based liquid crystal panel AOI inspection system according to claim 6, characterized in that: Analyze the abnormal surface: The brightness deviation values ​​of adjacent abnormal points in the abnormal surface are obtained. If the corresponding brightness deviation value exceeds the brightness deviation threshold, and each adjacent abnormal point constructs a nonlinear abnormal surface, a material mismatch signal is generated; the LCD panel detection platform sends a maintenance instruction and arranges the LCD panel design administrator to conduct an inspection; If the corresponding brightness deviation value exceeds the brightness deviation threshold, and each adjacent abnormal point constructs a linear abnormal surface, a traceability management signal is generated, and the LCD panel inspection platform sends a traceability instruction to the production line and arranges the production line supervisor to conduct a traceability inspection.

8. The artificial intelligence-based liquid crystal panel AOI inspection system according to claim 1, characterized in that: The vibration impact data and light impact data are respectively the vibration frequency generated by the surrounding environment on the detection auxiliary device when the detection area image is collected during the LCD panel detection process, and the floating amount of the numerical deviation of the real-time ambient light intensity corresponding to the surrounding environment and the fill light source of the detection auxiliary device when the detection area image is collected during the LCD panel detection process.

9. The artificial intelligence-based liquid crystal panel AOI inspection system according to claim 8, characterized in that: If the vibration impact data exceeds the vibration frequency generation threshold, or the light impact data exceeds the light intensity value deviation floating threshold, then the test risk assessment during the LCD panel inspection process is determined to be abnormal; If the vibration impact data does not exceed the vibration frequency generation threshold, and the light impact data does not exceed the light intensity value deviation floating threshold, it is determined that the test risk assessment during the LCD panel inspection is normal.

Images