Study On Automatic Identification And Classification Of Agricultural Products Using Computer Vision

The population of China is very large, but the per capita infield area is far off the average level of the world and the agricultural production value is very low, too. One of the main methods to increase the agricultural production value is to improve the post disposal level of the farm products, above of all.bruise automatic identification of tomato. While went without saying, there is great advantage substituting the computer vision for person vision to inspect the farm produce automatically. The primary researching tasks are as follows:Mathematical models, which transfer the process of image to operation of matrix, were build to describe tomato images. After experimenting, annular incandescence lamp was chosen as lamp-house. With white background, the acquired images were studied. The distance was max between two-peak value in B gray level histogram. A set of computer vision hardware system which is use for bruise evaluation and classification of tomato was set up.(2)Four methods of lower-layer image processing technique were studied. B gray level histogram segment method was adopted to segment the tomato from the background. To eliminate all kinds of noise, the quickly median filtering was adopted to filter the tomato image.The histogram pureline transfer was put into effect in order to increasing image.The method of lower-layer image processing technique could satisfy the demand of bruise detection and classification .(3)Distriction increasing for detecting bruise image of tomato was builtTraining multilayer feedforward neural networks with BP for detecting tomato bruise and classification was built, and testing precision reached 90%.(4)The application software of tomato bruise automatic identification and classificationwas programmed with Visual C++ 6.0. Its functions include 'file', 'image collection','lower-layer processing of image', 'bruise detection', 'network training', and 'classification'. The software could show the finally results of identification and classification. The results of experiment showed that the accuracy of classification is morethan 92%.This research is of great significance both in elevating the research level of our country in this field and in promoting the application of computer vision technology in agriculture engineering.Postgraduate : Wang Shu WenMajor: Agricultural electrization and automatizationTutor : Prof .Zhang Chang-Ii...