On-line Monitoring System For Spraying Quality Of Office Seat Parts Based On Machine Vision

In the traditional office chair manufacturing,most enterprises still rely on manpower for industrial production.The degree of automation of the production line is not high.When there are many orders,the labor intensity of workers is high,and the quality and quantity cannot be guaranteed.Therefore,this paper studies the spraying quality monitoring of office seat parts.Through the image acquisition and processing of office seat parts after painting,the production line is reasonably optimized to increase productiveness and automation.Taking the office seat production line of Yongyi Furniture Co.,Ltd.as the background and the five-star foot of office seat as the research object,this paper studies the on-line spraying quality monitoring system of office seat parts based on machine vision.The main work is as follows:Firstly,a machine vision inspection system is proposed,and the overall inspection framework is determined.Select the camera,lens,light source and other hardware according to the detection requirements,an industrial camera protection device is designed to make the camera work under the condition of paint pollution.And the industrial robot model used is determined.Secondly,according to the characteristics of the five-star foot of the office seat,an algorithm process for getting the area need to be detected is designed.After the image is collected.It is preprocessed by median filtering and an image enhancement method.Then the Improved Sobel operator is used for edge detection,and the threshold segmentation and edge skeletonization are used for contour extraction.The extracted contour is filled to obtain an approximate region,and then the obtained region is refined by open operation to obtain the final region.Finally,the complete extraction of the region to be detected is completed.Finally,the spraying integrity of the five-star foot of the office seat is detected by detecting the average value and deviation of the gray value of the area to be detected and the number of pixels exceeding a certain gray value.Thirdly,the process after testing is designed.By calculating the center point of the area to be detected,matching with the template to get the position of the foot and rotating,the five-star foot is divided into 11 areas,and then the defect areas are extracted by threshold segmentation of the five-star foot.These areas are screened according to the characteristics of the row and column coordinates of the 11 areas to judge the position of the defect.Then plan the path of the manipulator according to the obtained area.Finally,the detection software is written with QT to realize the image acquisition,display and processing of the five-star foot of the office seat.The experiment shows that the detection accuracy of the whole detection system is 98% and the average detection time is 122.305 ms,which verifies the feasibility of the algorithm proposed in this paper.