Kullback–Leibler Divergence-Based Feature Selection Method for Image Texture Classification

Nowadays, with the Big Data phenomenon, the need of the usage of feature selection (FS) methods is critical. FS can assist to address photograph analysis, and in reality it has been gaining importance within a rapid few years ago, however also to relieve the computational burden required for extracting records from the snapshots. An exhaustive assessment and evaluation of the latest contributions of FS to the field of image analysis is necessary. By eliminating unimportant features from the original set, selection of features significantly enhances the efficiency of texture classification, which is a key factor in machine learning performance. By selecting a characteristic, one can achieve accuracy in database type while also accelerating the class rate. The primary goal of the endeavor is to choose the most substantial skills in the function set to carry out a particular undertaking. Kullback–Leibler (KL) divergence approach showed that accuracy in classification is proven with the use of the present methods when feature selection is applied to Feature extraction.