Abstract
Rice is one of the most widely cultivated crops throughout the world. Health, quality and production potential of rice plants often gets severely affected by several plant diseases. Brown spot and blast diseases are among the worst diseases that seriously affect rice production worldwide. Both of these diseases are characterised by appearance of distinct shaped lesions on the plant leaves. The efforts to control the disease include using fungicides, pesticides and other such chemicals. However, no image processing based technique has yet been proposed for determination of such plant diseases. In this research, the colour texture of 400 rice leaf images is analysed using a pattern recognition method and the results of the study for early and accurate detection of leaf spots are presented. If the suggested technology is embedded into camera phones, a farmer can detect the diseases much before these diseases show their destructive symptoms without any help from a plant pathologist.