A Prototype of Automatic Rice Variety Identification System using Artificial Intelligence Techniques

Abstract

Identification of rice varieties automatically is very exigent task in the field of image processing and that necessitates extensive knowledge of agricultural research and the use of computational artificial intelligence technology. The development of Automatic Rice Variety Identification System using machine as well as computer vision, machine learning and deep learning techniques is the need of hour to solve this problem. Automatic Rice Variety Identification System identifies the variety of the rice grains by comparing it with trained system. It is very challenging task due to distinct qualities of grains of numerous rice varieties available in the market and having unique inherent features. In this paper, a prototype of Automatic Rice Variety Identification System is proposed. In the proposed system, a Rice Image dataset has been prepared from different rice varieties grown in Punjab. The developed dataset contains a total number of 9692 Rice Images. The proposed system has been tested using KNN classifier resulting in 93% accuracy.