Abstract: To diagnose and identify maize diseases quickly, take preventive measures timely and improve the diagnostic level of maize diseases,this article introduced the image processing technictue and the I3P neural network algorithm into the identification and diagnosis of maize diseases. Experiments show that conclusions of analysis which are gained by using the disease recognition model of the image processing technique to identify the disease samples collected match real conclusions of the practical application field and meet the the practical application of the agricultural production. The technology provides an effective method for ensuring the yield and quality of maize.

Key words: Maize disease, Image processing technique, Mathematical morphology, Neural network algorithm