Research on Leafy Vegetable Recognition and Navigation Line Extraction Algorithm Based on Machine Vision

Abstract

Aiming at the problems of current agricultural image recognition and navigation line extraction technology that are susceptible to light conditions and poor timeliness, the key technologies of leafy vegetable image recognition and farmland navigation line extraction are studied under the overhead mode of image acquisition equipment. In order to reduce the adverse effects of natural illumination changes on image recognition and segmentation, the 2Cg-Cr-Cb Algorithm is used to segment the image in grayscale. And for the image of leafy vegetables that are planted and grown evenly in agriculture, an improved DBSCAN center point detection method is adopted to realize the extraction of the center point and the fitting of the navigation line. And for the image that fails to fit the navigation line, the navigation line is compensated.