Intelligent button spraying and defective product sorting device based on AI visual inspection

Abstract

The invention relates to an intelligent button spraying and defective product sorting device based on AI visual inspection, and belongs to the technical field of button production.The device comprises a control module, an equipment body and a detection module, the equipment body and the detection module are electrically connected with the control module, the equipment body comprises a conveying belt and a sorting machine, and the conveying belt is used for conveying buttons; the sorting machine is used for sorting the buttons on the conveying belt; the detection module is arranged opposite to the conveying belt, and the detection module is used for shooting real-time images of buttons on the conveying belt and uploading image data to the control module; the control module is used for identifying defective buttons according to the image data and instructing the sorting machine to sort out the defective buttons on the conveying belt; the control module is used for counting the number proportion of the defective buttons, judges whether the number proportion exceeds a proportion threshold value or not and instructs the conveying belt to reduce the running speed when the number proportion exceeds the proportion threshold value.

Description

Technical Field

[0001] This invention belongs to the field of button manufacturing technology, specifically relating to an intelligent button spraying and defective product sorting device based on AI visual inspection. Background Technology

[0002] Buttons are an important part of clothing, and in modern production, buttons are generally produced using automated methods.

[0003] In automated production, sorting is required, such as sorting defective products or different shaped buttons. Current technologies mostly rely on manual methods to pick out buttons of the same color and size. However, prolonged work can lead to eye fatigue and errors, resulting in low accuracy. To address this, Chinese patent CN219253339U discloses a button visual sorting component, including a box, a visual sorting unit, and an outflow unit. The box contains a conveying device, an inclined plate on the upper right side, and a flow channel at the top. A horizontal plate is installed on the lower right side of the housing; the vision sorting unit includes an arch plate, a sliding plate, an industrial camera, a cylinder, a vertical plate, a vertical channel, and a proximity sensor. An arch plate is installed on the right end of the flow channel, and a sliding plate is connected to the inside of the arch plate. An industrial camera is installed at the lower end of the sliding plate, and a cylinder is installed on the right end of the arch plate. The telescopic end of the cylinder is connected to the upper end of the vertical plate. It uses an industrial camera to perform visual inspection of color and other specifications, making it less prone to errors when selecting buttons and achieving a higher selection accuracy. After selection, buttons of different colors and specifications can be sorted and discharged through the outflow unit without manual operation, resulting in higher button sorting efficiency.

[0004] In conventional solutions, button defects are often small due to their size, requiring high-resolution visual inspection for extended periods. This necessitates a balance between accuracy and speed. Insufficient speed hinders mass production, while high speed can lead to higher false negative rates. The aforementioned solutions do not address these issues. Therefore, a smart button coating and defect sorting device based on AI visual inspection is needed that balances production efficiency with defect recognition accuracy. Summary of the Invention

[0005] To address the aforementioned problems in the existing technology, this invention provides an AI-based intelligent button spraying and defective product sorting device, which balances production efficiency with ensuring accurate defect identification.

[0006] The objective of this invention can be achieved through the following technical solutions: A smart button spraying and defective product sorting device based on AI visual inspection includes a control module and an equipment body and a detection module electrically connected to the control module. The equipment body includes a conveyor belt and a sorting machine. The conveyor belt is used to transport buttons, and the sorting machine is used to sort the buttons on the conveyor belt. The detection module is positioned directly opposite the conveyor belt. The detection module is used to capture real-time images of the buttons on the conveyor belt and upload the image data to the control module. The control module is used to identify defective buttons based on image data and instruct the sorting machine to sort out the defective buttons on the conveyor belt. The control module is used to count the percentage of defective buttons. The control module determines whether the percentage exceeds a threshold and instructs the conveyor belt to reduce its running speed when the percentage exceeds the threshold.

[0007] As a preferred embodiment of the present invention, the control module is used to count the proportion of defective buttons Z, and the control module is used to instruct the speed of the conveyor belt to be adjusted to V, where V=V0×(1-Z / Z0)×c, V0 is the pre-inputted reference speed of the conveyor belt, Z0 is the pre-inputted proportion threshold, and c is the pre-inputted correction coefficient.

[0008] As a preferred embodiment of the present invention, the control module is pre-set with several speed change points, and when the control module instructs the conveyor belt to adjust its speed, the execution process is adjusted in segments according to the preset speed change points.

[0009] As a preferred embodiment of the present invention, the control module adjusts the number of speed change points to n, where n = Z / Z0 × d, and d is a pre-input constant.

[0010] As a preferred technical solution of the present invention, it further includes an environmental monitoring module, which is used to collect the surrounding visibility and upload it to the control module. The control module is used to calculate the comprehensive confidence coefficient based on the visibility and determine whether the comprehensive confidence coefficient is lower than the confidence threshold. When the determination result is yes, the control module increases the image acquisition frame rate of the detection module based on the comprehensive confidence coefficient and the conveyor belt speed.

[0011] As a preferred technical solution of the present invention, the environmental monitoring module is used to collect the surrounding visibility N and upload it to the control module. The control module is used to calculate the comprehensive confidence coefficient C based on the visibility N and increase the image acquisition frame rate to A1 times the original value, where C=N0 / N×k, N0 is the standard visibility value, k is the pre-input environmental correction factor, and A1=V / V0×(C / C0)×e, where C0 is the pre-input confidence threshold.

[0012] As a preferred embodiment of the present invention, it further includes a control panel, which is used to input the values ​​of C0, V0, Z0, d, e and k.

[0013] The beneficial effects of this invention are as follows: (1) By setting the control module to count the percentage of defective buttons, and judging whether the percentage exceeds the threshold, the conveyor belt is instructed to reduce its running speed. When the defect rate increases, it is likely that there is an abnormality in the spraying process or an abnormality in the detection. The production speed is automatically reduced to reserve more detection time, so as to balance production efficiency and ensure the accuracy of defect identification. (2) By pre-setting several speed change points, when the instruction conveyor belt adjusts its speed, the execution process will be adjusted in segments according to the preset speed change points to achieve smooth acceleration change, avoid the impact of sudden speed change on the mechanical structure of the conveyor belt, and at the same time avoid the displacement of buttons on the conveyor belt due to inertia caused by sudden production speed change, as well as the rhythm disorder caused by other processes in the production line being unable to keep up with the speed change. (3) By adjusting the number of speed change points to n by the control module, where n=Z / Z0×d, and d is a pre-input constant, the number of speed change points is dynamically adjusted according to the current defect ratio Z, ensuring that the precision of acceleration smooth adjustment matches the actual needs. When the defective button ratio is high and the requirement for detection accuracy is higher, efficiency is sacrificed to increase the number of speed change points to achieve more precise speed adjustment, further reduce the displacement probability, and improve detection stability. When the ratio is low, the number of speed change points is reduced to improve adjustment efficiency and avoid unnecessary delays. (4) By setting the environmental monitoring module to collect the visibility around the collection, calculate the comprehensive confidence coefficient, and increase the image acquisition frame rate of the detection module according to the comprehensive confidence coefficient and the conveyor belt speed. When the environmental factors have a large interference and the detection accuracy may decrease, the acquisition frame rate is increased to improve the image quality. When the environmental factors have a small interference and the detection accuracy is guaranteed, the frame rate is appropriately reduced to save computing resources. It is also suitable for situations where the production speed is fast due to the low proportion of quantity. Attached Figure Description

[0014] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.

[0015] Figure 1 This is a block diagram of the control loop of the present invention. Detailed Implementation

[0016] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided.

[0017] Please see Figure 1 A smart button spraying and defective product sorting device based on AI visual detection includes a control module and an equipment body and a detection module electrically connected to the control module. The equipment body includes a conveyor belt and a sorting machine. The conveyor belt is used to transport buttons, and the sorting machine is used to sort the buttons on the conveyor belt. The detection module is positioned directly opposite the conveyor belt. The detection module is used to capture real-time images of the buttons on the conveyor belt and upload the image data to the control module. The control module is used to identify defective buttons based on image data and instruct the sorting machine to sort out the defective buttons on the conveyor belt. The control module is used to count the percentage of defective buttons. The control module determines whether the percentage exceeds a percentage threshold and instructs the conveyor belt to reduce its running speed when the percentage exceeds the threshold. When the quantity exceeds the threshold, it indicates that the defect rate is too high, which may indicate that there is an abnormality in the production line or that the equipment is unstable. At this time, reducing the conveyor belt speed can reduce the detection omissions and misjudgments caused by high-speed operation, and at the same time, it can buy adjustment time for subsequent processes and avoid the continuous accumulation of defective products. By setting the control module to count the percentage of defective buttons, and determining whether the percentage exceeds a threshold, the conveyor belt is instructed to reduce its running speed. When the defect rate increases, indicating a possible abnormality in the spraying process or detection, the production speed is automatically reduced to allow more inspection time, thus balancing production efficiency with the accuracy of defect identification.

[0018] In the above adjustment process, specifically, the control module is used to count the proportion of defective buttons Z, and the control module is used to instruct the speed of the conveyor belt to be adjusted to V, where V=V0×(1-Z / Z0)×c, V0 is the pre-input reference speed of the conveyor belt, Z0 is the pre-input proportion threshold, and c is the pre-input correction coefficient.

[0019] During the speed adjustment process, if the speed adjustment is too fast, it may cause the system response to lag and impact the mechanical structure of the conveyor belt. Therefore, it is necessary to apply a smoothing treatment to the speed adjustment process and adopt a segmented adjustment strategy. For this purpose, the control module has several speed change points set in advance. When the control module instructs the conveyor belt to adjust the speed, the execution process will be adjusted in segments according to the preset speed change points. There is a fixed time interval between each segment adjustment to ensure that the speed change is smooth. For example, there are five speed change points. When the control module commands the speed to be adjusted from V to 0.8V, the control module divides the speed adjustment process into five segments, each segment reducing the original speed by 4%, with a 2-second interval between adjacent segments to ensure that the conveyor belt decelerates smoothly. By pre-setting several speed change points, the conveyor belt will adjust its speed in segments according to the preset speed change points, achieving smooth acceleration changes. This avoids the impact of sudden speed changes on the conveyor belt's mechanical structure, and also prevents buttons on the conveyor belt from shifting due to inertia caused by sudden changes in production speed, as well as the disruption of the production line's rhythm caused by other processes being unable to keep up with the speed changes.

[0020] Because buttons are lightweight and have low friction with the conveyor belt, the number of speed change points required varies depending on the situation. When the defect rate fluctuates slightly, fewer speed change points can be used to achieve a quick response, while when the defect rate exceeds the standard significantly, more speed change points are used for fine adjustment to ensure a smooth and reliable deceleration process. Therefore, the control module adjusts the number of shift points to n, where n = Z / Z0 × d, and d is a pre-input constant; By adjusting the number of speed change points in the control module to n=Z / Z0×d, the number of speed change points is dynamically adjusted according to the current defect ratio Z. This ensures that the precision of acceleration adjustment matches the actual needs. When the defective button ratio is high and the requirement for detection accuracy is higher, efficiency is sacrificed to increase the number of speed change points to achieve more precise speed adjustment, further reducing the probability of displacement and improving detection stability. When the ratio is low, the number of speed change points is reduced to improve adjustment efficiency and avoid unnecessary delays. In addition, the environment can also affect monitoring. For example, changes in lighting and dust concentration can interfere with the recognition effect of the visual inspection system. Compensation settings are needed for this factor. Therefore, an environmental monitoring module is also included. The environmental monitoring module is used to collect the surrounding visibility and upload it to the control module. The control module is used to calculate the comprehensive confidence coefficient based on the visibility and determine whether the comprehensive confidence coefficient is lower than the confidence threshold. When the determination result is yes, the control module increases the image acquisition frame rate of the detection module based on the comprehensive confidence coefficient and the conveyor belt speed. Specifically, the environmental monitoring module is used to collect the surrounding visibility N and upload it to the control module. The control module is used to calculate the comprehensive confidence coefficient C based on the visibility N and increase the image acquisition frame rate to A1 times the original value. , The standard visibility value is given by k, which is a pre-input environmental correction factor. A1 = V / V0 × (C / C0) × e, where C0 is a pre-input confidence threshold. When N is large, visibility is good, C approaches a smaller value, and A1 is close to 1, so the image acquisition frame rate remains at the baseline level. When N decreases, i.e. when the environmental visibility decreases, C increases accordingly, A1 is greater than 1, and the image acquisition frame rate is increased accordingly to increase the sampling density per unit time. Similarly, when V is large, i.e. the conveyor belt speed is fast, in order to ensure that sufficient image information can still be obtained under low visibility or high dynamic conditions, A1 increases with V, thereby increasing the frame rate to compensate for the imaging blur and missed detection risk caused by motion. By setting up an environmental monitoring module to monitor the visibility around the data acquisition point and calculating the comprehensive confidence coefficient, the image acquisition frame rate of the detection module is increased based on the comprehensive confidence coefficient and the conveyor belt speed. This allows for improved image quality by increasing the acquisition frame rate when environmental factors cause significant interference and detection accuracy may decrease, and for saving computing resources when environmental factors cause less interference and detection accuracy is guaranteed. It also adapts to situations where the production speed is faster due to a lower quantity.

[0021] To facilitate the input of various constants, a control panel is also included. The control panel is used to input the values ​​of C0, V0, Z0, d, e, and k. The control panel includes at least one display that is electrically connected to the control module.

[0022] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A smart button spraying and defective product sorting device based on AI visual inspection, characterized in that: The device includes a control module, a device body electrically connected to the control module, and a detection module. The device body includes a conveyor belt and a sorting machine. The conveyor belt is used to transport buttons, and the sorting machine is used to sort the buttons on the conveyor belt. The detection module is positioned directly opposite the conveyor belt. The detection module is used to capture real-time images of the buttons on the conveyor belt and upload the image data to the control module. The control module is used to identify defective buttons based on image data and instruct the sorting machine to sort out the defective buttons on the conveyor belt. The control module is used to count the percentage of defective buttons. The control module determines whether the percentage exceeds a threshold and instructs the conveyor belt to reduce its running speed when the percentage exceeds the threshold.

2. The intelligent button spraying and defective product sorting device based on AI visual inspection according to claim 1, characterized in that: The control module is used to count the percentage of defective buttons Z, and the control module is used to instruct the speed of the conveyor belt to be adjusted to V, where V=V0×(1-Z / Z0)×c, V0 is the pre-inputted reference speed of the conveyor belt, Z0 is the pre-inputted percentage threshold, and c is the pre-inputted correction coefficient.

3. The intelligent button spraying and defective product sorting device based on AI visual inspection according to claim 2, characterized in that: The control module has several speed change points preset. When the control module instructs the conveyor belt to adjust its speed, the execution process is adjusted in segments according to the preset speed change points.

4. The intelligent button spraying and defective product sorting device based on AI visual inspection according to claim 3, characterized in that: The control module adjusts the number of speed change points to n, where n = Z / Z0 × d, and d is a pre-input constant.

5. The intelligent button spraying and defective product sorting device based on AI visual inspection according to claim 4, characterized in that: It also includes an environmental monitoring module, which collects the surrounding visibility and uploads it to the control module. The control module calculates a comprehensive confidence coefficient based on the visibility and determines whether the comprehensive confidence coefficient is lower than a confidence threshold. When the determination result is yes, the control module increases the image acquisition frame rate of the detection module based on the comprehensive confidence coefficient and the conveyor belt speed.

6. The intelligent button spraying and defective product sorting device based on AI visual inspection according to claim 5, characterized in that: The environmental monitoring module is used to collect the surrounding visibility N and upload it to the control module. The control module is used to calculate the comprehensive confidence coefficient C based on the visibility N and increase the image acquisition frame rate to A1 times the original value, where C=N0 / N×k, N0 is the standard visibility value, k is the pre-input environmental correction factor, and A1=V / V0×(C / C0)×e, where C0 is the pre-input confidence threshold.

7. The intelligent button spraying and defective product sorting device based on AI visual inspection according to claim 6, characterized in that: It also includes a control panel for inputting the values ​​of C0, V0, Z0, d, e, and k.

Images