India has a huge population and the major source of food supply is from agriculture. Agricultural lands are getting destroyed mostly due to crop diseases, pests and plant diseases. Nowadays it increased fiercely. Plant pathogens are fungi, bacteria, virus or nematodes that damage plant parts such as leaf, panicle, node, stem and roots. So In agriculture, one of the latest researches is detection and classification of crop diseases using images of leaves from plants. The detection of plant diseases by using the image processing methods can help the farmers to protect the agricultural field from getting destroyed or affected. Paddy is a staple crop for much of the world's population. In this research we discuss classification and detection of paddy Leaf diseases using convolutional neural network. We captured paddy leaf images from the field for normal, sheath rot, rice blast, bacterial leaf blight, rice blast, brown spot, rymv and rice tungro for processing the image. In pre-processing, background exclusion is done based on the hue value to extract the non-diseased and diseased part. Feature Extraction is executed by applying Convolutional Neural Network. The proposed system achieved a high accuracy.