Chapter 14 - Machine learning techniques for agricultural image recognition

Abstract

Image processing is a promising analyzing tool in the farming sector. As a result of assessing based on appearance features, which can eliminate additional steps such as expert analysis and costly and time-consuming experiments, agriculture takes the advantage of image recognition. Image-processing strategies due to their soundness properties such as being accurate, quick, and cost-sensitive have recently been widely and increasingly used in the agricultural industry. The purpose of this study is to focus on the use of image processing in the agricultural field. This chapter is composed of three main parts. First, the steps of image analysis are briefly reviewed. Next, common machine learning techniques for image recognition used in agriculture are introduced and analyzed. Finally, practical applications of image processing in agriculture’s tasks including soil assessment, irrigation, leaf analysis, weed detection, pest control, disease recognition, vegetation measurement, monitoring plant growing, and fruit/food grading are highlighted.