Study on the method of visual navigation baseline identification and extraction of agricultural machinery

In this paper a method is presented to extracts visual navigation baseline based on the processing image of farmland. Firstly, the weighted average method(2G-R-B) is used for processing original image. And improved Otsu is used to segment the gray scale image. Then the binary image of the division is denoised with small area method. The center points are identified of the denoised image, and points´ position are determined. Based on the distribution of the center point and the improved Hough transform, the invalid and singular points are deleted effectively. The navigation baseline is generated by using least squares. The conclusion is that the method is effective and has the real-time property.