A Review: Plant's Leaf Analysis based on Normalized Features Set using Image Processing

Leaf image patterns are the primary characteristics of a plant under the respective category. Pattern image features can be mapped to the respective plant category when made input to a SVM or ANN classifier. Pattern image features in different domains including radial, color, texture, morphological, statistical and ratio domains are extracted out. Image pattern features normalization with respect to size and rotation is an important aspect when made input to classifier. Mean radius is used to normalize the image features with respect to size. Largest chord of the leaf image is approximated to its mid rib, and, rotation aspect is covered by orienting the leaf image with respect to the largest chord i. e. mid rib. Computation of features like radii, area, perimeter and standard deviations in different quadrants contributes fair resolution at close level. Support vector machine classifier categorizes the features set into respective plants class to an equal degree of high accuracy.