Date of Publication :30th March 2018

Abstract: Image processing is a method to perform some operations on an image in order to extract some useful information. Feature Extraction is one of the significant techniques in image processing. In this Paper, the analysis was processed in three steps. First, the leaf images were preprocessed using Median Filter and transformed into LBP (Local Binary Pattern).Second, the retrieval was performed using feature extraction methods such as GLCM (Gray Level Co-occurrence Matrix), CGLCM (Color Gray Level Co-occurrence Matrix), Shape and Hybrid. In each method, various statistical measurements like Contrast, Correlation, Energy, Homogeneity, Entropy, Area, Perimeter, and Eccentricity were calculated. The KNN classification algorithm was used to retrieve the relevant images. The dataset was collected by capturing some 10 types of leaves using Digital Camera. Each type contains 10 images and used for analysis. Finally, the Performance measurement such as Precision and Recall were measured for each extraction method. The Hybrid extraction method produces the better result than the individual method. This analysis is used to identify the medicinal leaves and its use from the large collection of Database

Reference :