Research On Algorithm Of Rice Color Sorting Based On Grey

With the continuous developing of the level of socio-economic and the living standard, the higher accuracy to the grain processing is required. The features of the advanced color sorter, domestic or abroad, are learned, and the algorithm of color election is designed in this thesis.This thesis analyses and summarizes the development and the current research of the color sorter firstly, both domestic and abroad. Some relevant knowledge of the color election, such as the algorithm of the image denoising, edge detection and the threshold selection in image processing are introduced. The pretreatment such as gray processing and denoising is discussed in this thesis.There are four sections in this thesis to talk about the details of the design to the color election algorithm. They are the uniform standards for data, how to determine the edge detection, the method to produce the threshold and how to determine the discrimination rules of the color election. The return tracking to the threshold is implemented based on statisical threshold method of 3σ. The information of the threshold is enriched in the premise that the test sample is limited. The method of the color election discrimination is selected and the background is updated, because the artificial operation reduced and the efficiency is improved in the process of color election.Simulation results shown that the algorithm has better effect of the discrimination of the distinct points on rice.