Research On The Control System Of Wolfberry Sorting Machine Based On Machine Vision

Wolfberry has covered 850,000 mu cultivating areas in Ningxia, which has reached more than 45% of nationwide wolfberry cultivating areas. To maximize the selling price and business value, it is necessary to grade the wolfberry. However, to date grading work of wolfberry is mainly manual operation which means high labor cost, low accuracy and efficiency. Besides, it is difficult to guarantee food safety because dried fruit of wolfberry has direct contact with workers. The difficulties of grading work of wolfberry have limited the development of wolfberry industry to some extent.Up to now, it is Wolfberry Color Sorter the main existing equipment of grading wolfberry. Wolfberry Color Sorter takes pictures of wolfberries when they move into the image acquisition under a certain rule by CCD camera. After computer processing, shaping and amplifying circuit starts the shutter by controlling electromagnetic and then blows wolfberries with different colors into different pipes. However, grading indexes are not only refer to color, but also size, shape,bruise,etc. in practical production. With the development of computer technology and intelligent instrument, plenty of applications of automation equipment of computer, electronics and mechanics in agricultural production make it possible to research and development equipment of grading wolfberry with more grading indexes. My essay does research on detection algorithm to check the size,shape, color and bruise of wolfberry by using computer vision and embedded technology; Based on detection algorithm, my task is to design a circuit system on the core of embedded MCU and then to apply it to wolfberry sorter. Thus it can fulfill its function of detection and grading online and real-time. At the same time, it can offer certain practice experience for the development of detection and grading equipment.My essay focuses on:(1)Preprocess the images by researching proper algorithm considering graying, denoising, edge detection and ellipse fitting of images. Do research on detection algorithms of each index of the quality of wolfberry.(2)Design circuit of wolfberry sorter on the core of STM32F429 which can acquire images, identify characteristics and control in grades.(3)Fulfill the design and operation of software of wolfberry sorter by transplanting the detection software to the embedded system according to the study of detection algorithm of quality of wolfberry.The test results suggest that the wolfberry detection algorithm which my essay focused on can detect the size, shape, color and bruise accurately. Wolfberry Sorter based on embedded technology can grade wolfberry by its size and shape well.