Guidance Line Identification for Agricultural Mobile Robot Based on Machine Vision

Abstract

Conventional navigation path detection algorithms based on machine vision are difficult to guarantee the guidance line accuracy, real-time and anti-interference. This paper carries out research on visual guidance line recognition for monocular vision agricultural mobile robot under natural environment. According to the characteristics of the crop rows in the image, a method of crop lines identification based on improved genetic algorithm was proposed. To approximate the trend of a crop row in the image to a line, two points from image bottom and top side were randomly selected to code as a chromosome. Through multiple genetic evolutions, the chromosome with the highest fitness value is selected as a crop center line and the guidance line can be obtained according to the two adjacent crop lines. Experimental results showed that, the guidance line recognition method based on improved genetic algorithm can quickly and accurately detect the navigation line. Furthermore, it not only has good fitness for different crops but also has nicer adaptability for different growth stages of crop in the farmland.