Agriculture is the main component of the Indian economy. Preparation of soil, sowing, pest control, adding manure and fertilizers, irrigation, and harvesting are major processes involved in agriculture practices. Pest control is the most challenging task faced by farmers. Hence the system which being utilized the modern image processing technique is proposed to identify the infectious plant's diseases and the same is reported in this monograph. In this work, a recognition system is developed which capable of identifying the type of diseases affected by plants using images of their leaves. This is done by comparing the captured image of leaves with an image of healthy leaves. The recognition system runs the pre-processing and feature extraction techniques to identify diseases caused by pests. Here, a canny algorithm is used to identify and analyze diseases in plants since it is effective than others. In this work, the banana tree and paddy are considered for the analysis, and results are presented. It is the evidence that the canny algorithm gives an exact disease identification with vital accuracy and hence proposed system improves the cultivation rate.