Research On The Moving Food Machine Vision Identification And Location Technology

The rapid development of the society brought qualitative leap to the people’s standard of living, along with the improvement of people’s living quality requirements, food safety and hygiene to get people’s attention, food industry restructuring and upgrading become the consensus of the industry. The machine vision and high-speed parallel robot is introduced into the traditional food production industry, the use of visual system for real-time detection and tracking of food production line, using high-speed robot for fast sorting, has become the development trend of modern food production industry.Aiming at the actual demand for food production lines, based on machine vision and the high-speed sorting robotics, building a machine vision positioning systems, positioning system based on machine vision task, select the key components, and analyzed the working principle of automatic sorting food high speed parallel robot, the machine vision positioning algorithm of the moving food on the conveyor belt is proposed.According to the needs of the moving food machine vision positioning system, the geometric feature extraction algorithm of the moving food is studied. For circular food, using the extreme value point average algorithm to determine the center characteristics of the food; For rectangular food, to determine its position with the improved corner detection method; For oval food,using the improved least square method to determine the center and the rotation Angle; For other non-Regular Graphs, adopt the general image processing algorithm to extract geometric characteristic, Through the programming algorithm to complete the description of food pose. The process of coordinate transformation in machine vision system are introduced, image coordinates is obtained by experiment and coordinate the changing relationship between the manipulator.In order to facilitate the extraction of geometric features of the image, research on the image preprocessing method. Firstly, image enhancement, mainly used for the two value processing. Then remove the noise, image smoothing, Finally, through the experimental comparison of several common edge detection method, selects the Canny operator edge detection, get a clear edge character, provides important data for later of geometric feature extraction.In the VC++2008integrated development environment, based on the professional machine vision software Sherlock, the machine vision positioning software is developed. The software based on of Sherlock’s image acquisition and image enhancement functions, the moving food through two value processing, highlight the edge features, through the corresponding movement food geometric feature extraction algorithm, obtain position of food on conveyor belt, to realize the automatic positioning and sorting of food.