PLANT LEAF DISEASE IMAGE DETECTION AND CLASSIFICATION USING ANN

Abstract : Digital media archives are increasing to colossal proportions in the world today, which includes audio, video and images An Image refers as a picture produced on an electronic display .A digital image is a numeric representation of a two-dimensional image. Digital image processing refers to processing of digital images by using digital computers. Nowadays, most of the applications prefer digitalized version, to reduce memory space. Lot of application depends on digital images. One of the important application is medical image processing.

 EXISTING SYSTEM :

 The Feature Discontinuity extracts the regions having different properties like intensity, color, texture etc. Similarity groups the image pixels into groups with some predefined criteria. PCA Based on pixel similarity with the neighboring pixel, the algorithm used is region based. In leaf disease identification, segmentation is used to identify the diseased area. From this, features of a region are computed; we have to extract the features corresponding to the disease in this system Not Clearly Recognition Leaf Disease Result.

 DISADVANTAGE :

 a. Noise High b. Output Not Clearly Recognition Leaf Disease c. To identify the disease in plant is less accurate

 PROPOSED SYSTEM :

 In the proposed work, we have concentrated on identification of Leaf Spot disease and Leaf Miner from the photographic signs and classify them using image processing techniques. The proposed framework has been implemented in three steps. First, image segmentation is performed using K means clustering to identify the infected area. In the next step leaf features are extracted from segmented regions using feature extraction techniques such as GLCM. These features are then used for classification into infected or non-infected leaf type. As third step these features given to the classifier to classify the disease in the cotton crop. We used ANN classifier to obtain efficient results.

 ADVANTAGE :

 K-means clustering is a method of vector quantization, originally from signal processing, that is popular for cluster analysis in data mining. K-means clustering aims to partition n observations into k clusters in which each observation belongs to the cluster with the nearest mean, serving as a prototype of the cluster.