An Enhanced Transformed Encoded Local Binary Pattern for Feature Extraction and Integrated Tabu Search with SDS based Feature Selection

Abstract

The majority of the Indian economy depend on agriculture, so distinguishing any diseases crop in beginning phases is extremely essential as these diseases in plants causes an enormous drop in the creation and economy of the farmers and accordingly, debasement of the banana crop which underscore on the early recognition of the plant stem disease. Nowadays, location of banana stem diseases has turned into an intriguing issue in the space of interest of the researchers. Farmers followed a customary methodology for recognizing and identifying diseases in plants with naked eyes, which didn't assist much as the disease with having made a lot of harm the plant. For banana stem feature extraction we proposed Enhanced Transformed Encoded Local Binary Pattern Algorithm (ETELBA). Choosing an optimal subset from unique enormous feature set in the plan of pattern classifier is a significant and troublesome issue. In this paper, we proposed Integrated Tabu and Stochastic Diffusion Search Algorithm (ITSDS) to tackle this feature selection issue and contrast it and classic algorithms.