ISSN 2348-1196 (print) International Journal of Computer Science and Information Technology Research ISSN 2348-120X (online) Vol. 3, Issue 1, pp: (63-69), Month: January - March 2015, Available at: www.researchpublish.com Page | 63 Research Publish Journals An Assessment on Detection of Plant Leaf Diseases and Its Severity Using Image Segmentation 1 K.Muthukannan, 2 P.Latha, 3 P.Nisha, 4 R.Pon Selvi 1 Associate Professor, Einstein college of Engineering, Tirunelveli, India 2 Associate Professor, Government college of Engineering, Tirunelveli, India 3, 4 PG Student, Einstein college of Engineering, Tirunelveli, India Abstract: Diseased plant leaf image segmentation is a basic pre-processing task to separate leafs from the background and detects the disease affected portion. Although many methods are proposed, it is still difficult to accurately segment the affected portion and to measure the severity of disease in a leaf image. In this survey paper, some segmentation techniques are illustrated to segment the spots of lesion regions in leaves and the measurement of severity of disease affected in a leaf. Keywords: detection, segmentation, colour edge extraction, hybrid techniques, severity measurement. I. INTRODUCTION Plants exist everywhere in this world, detecting the plant leaf diseases plays a vital role in the agricultural field. Plant diseases degrade the quantity and quality of agricultural products. Diseases, insects and pests are the major problems that threaten any plant cultivation, which leads to heavy loss in production. Plant leaf disease detection is important research topic, which automatically detects the causes of diseases in unhealthy leaves as soon as they appear. In this survey paper the diseased leaf segmentation techniques and severity measurement techniques are described. The main step in image processing is image segmentation, where the image is subdivided into number of meaningful segments. The segmented parts give some information in the form of colour, texture or intensity. And also it’s importan t to extract the boundaries of an image. Still there is no perfect method for image segmentation because each image has its own properties. Some the famous image segmentation is Edge based segmentation, Fuzzy theory based segmentation, Artificial Neural Network (ANN) based segmentation, threshold based image segmentation, and Region based image segmentation. This paper mainly focus on some hybrid segmentation techniques such as discrete wavelet transform and k- means clustering, colour edge extraction and seeded region growing technique, image edge detection technique and threshold based segmentation. And plant leaf diseases can be measured in many ways based on intensity, incidence, prevalence and disease severity. Disease severity can be measured by the absolute area of the sampling unit which shows the symptoms of a disease. Today there is no any standard to calculate the severity of a disease affected leaf image. The steps involved in severity measurement include image acquisition, diseased leaf area calculation, total leaf area calculation and calculation of disease severity. This is used for early prediction of plant leaf diseases and reduces the loss in agricultural productions by using appropriate insecticides and to increase the crop yields. II. REVIEW OF LITERATURE Evaluation of Cotton Leaf Spot Disease Detection Using [1] The paper titled Advance Computing Enrichment Image Edge Detection by P.Revathi and M.Hemalatha describes about the image edge detection segmentation techniques to segment the disease affected cotton leaf. Here Image analysis purposes are to detect the diseased leaf, to measure the affected area by disease, to find the boundary of the affected area, to determine the colour of the affected area, to identify the diseased spot correctly. Here they have used image edge detection segmentation and advance computing techniques.