Extracting the navigation path of an agricultural plant protection robot based on machine vision

Abstract

Intelligent agricultural machinery is the development trend of future agricultural machinery. The automatic navigation of agricultural machinery is one of the core technologies in precision agriculture, which is widely used in the production processes of farming, seeding, fertilization, spraying, and harvesting. To achieve autonomous operation, it is necessary to develop a reliable and stable navigation system. The navigation route detection algorithm based on computer vision is the core of the automatic navigation system. Aiming at the extraction of visual navigation lines, this paper proposes a method that combines contour computing with vertical projection. It uses a position clustering and the shortest distance algorithm to fit the navigation line through the ordinary least square method. The navigation parameters are calculated from the navigation line and provided to the agricultural robot. In the experiments, it is shown that the detection of multiple rows of crops and autonomous walking along the rows can be achieved under indoor illumination. As the result, the maximum navigation path deviation is less than 0.5°, and the average processing time of each loop is about 25ms, which verifies the high accuracy and efficiency of the proposed algorithm, and proves that it can be extended to practical agricultural robots.