Agricultural Robot Navigation Path Recognition Based on K-means Algorithm for Large-Scale Image Segmentation

In order to solve the problem of agricultural robot navigation path recognition in the uneven illumination and complex background environment which lead to the poor accuracy of navigation path, a clustering algorithm for image segmentation is used in this paper. By introducing the Lab color space and K-means algorithm, the K-means clustering process can be performed with large-scale segmentation of the region of interest in the image. After clustered twice, the image can separate the path information of the farmland from background. The navigation path can be fitted by using the linear least squares method. For illustration, an image of the medlar farmland line is utilized to show the feasibility of this method. Experience results show that the method of clustering and segmenting the region of interest based on K-means algorithm can effectively improve the accuracy of image segmentation and solve the influence of uneven illumination and complex background environment on farmland navigation path accuracy.