Computer Vision Detection And Granding System Used In Apple Automatic Production And Packing Line

With the development of computer technology and industrial production, online automated detection of production, has the important application prospect in the modernized automatic industrial production. Using computer vision substituted the vision of human, has many advantages in automatic production line of products for automatic detection , such as real-time, objective, harmless. At such a background, apple was chosen for study to satisfy the demand for related businesses. Based on computer vision of the automatic detection methods and techniques were discussed. The main research contents are presented as follows:(1) Image acquisition, a mathematical model was built by the apple relative to the time required for the establishment of the steady-state, by which the best position, angle and volume of the camera was calculated to image acquisition. 3 CCD position parameters mathematical model was established, so that the images of ensuring productivity and accuracy were acquired. A derived center self-calibration method was used to demarcate the system.(2) A vector improved median filtering method was applied to filtering processing, and Z-transform was applied to image restoration of the motion-blurred image restoration,. Then "a fast edge detection method using templates" was used on the edge contour detection.(3) A new arithmetic is proposed to inspect the dimension of apple. According to the shape of apple, a three-dimensional cube-shaped model was built for elementary fruit-shaped detection.(4) Law of defects divided was proposed, analyzed the color ratio between the skin and defect of apple, and which is successfully used to segment the surface defects of fruits. A logical way was rapidly proposed to distinguish defect and stem, and which was satisfied with the results. In this paper, a considered curvature factor, "rapid transformation of pixels method by Calculus" based on pixel transform was proposed accurately to calculate the defects area, and which got a very good results.Application program development has been done based on the WindowsXP operating system and in application of MATLAB programming language. The result comes to the anticipant. In this paper, the results of research on the quality detection of fruits and vegetables were important to the promotion and practical value.