The Research On Image Processing Methods Of Identification Weeds And Soil Background

Weeds pose one of the most important threats to our supplies of food and fiber. Losses in both yield and quality of crops due to weeds, as well as costs of weed control, constitute an enormous problem in all agricultural areas. The research on weeds detection and control is very significant for elimiating the weeds and scientific agriculture. A color machine vision system is developed to identify weeds and soil background. The hardware system consists of a color canon camera, a MICROTER E6 scanner, a Pentium 586/166 computer and a 14 inch color display. The method of using the HSI(Hue, Saturation, and Intensity)color system proved highly effective for color evaluation and image processing. The hue color feature is used and transformed into gray scale. Based on the experiments, The class clustering method, K-Mean, is used to sort the optimal threshold of the gray- hue histogram. In the processing, mean filter, quickly- median filter, and morphology filter are used to erase the noise. The Robert operator is selected to detect the edge of the weeds. The area and centroid of the weed are computered. Several weeds (Japanese Hop, Lambsquarters, Sweet Wormwood, Chingma Abutilon, Ascendent Crabgrass, Horseweed, Redroot amaranth)commonly found in the wheat fields are used to test the system. The results showed that the machine vision system is effective for different sun shining weed images. The successful rate is 92%.