Grape leaf segmentation for disease identification through adaptive Snake algorithm model

Abstract

Fruits are the eminent export agriculture product for any country, especially grape, it is used for making wines and developing raisins. Moreover, Viticulture has proven to be one of the highly profitable industry from the economic point of view. However, in viticulture grape quality should be of top-notch quality, moreover throughout the research it is observed that grape quality is degraded mainly due to the plant disease. In past several researcher have tried their efforts to detect the early identification of disease. Segmentation plays bigger role identifying the disease, Hence in this paper we propose an Adaptive Snake Model for segmentation and region identification. ASA (Adaptive Snake Model) is two phase segmentation model namely common segmentation and absolute segmentation. Through common segmentation, we achieve the fast segmentation and through the absolute segmentation, we achieve the better accuracy. Moreover for evaluation of Adaptive Snake Algorithm two standard dataset i.e. PlantLevel and PlantVillage dataset, for further evaluation we have compared with various state-of-art technique in terms of various performance metric such as PSNR, Dice, Manhattan, Recall and Jaccard. Adaptive Snake Algorithm performs better than the other existing methodology.