Support Vector Machine-based image segmentation approach for automatic agriculture vehicle

Support Vector Machine, a statistic procedure, is robust and has good performance when applied to farmland image segmentation. It can effectively identify the crop rows, even though they have some intersection at some points because of weeds of the leaves growth of the crops. However, the Support Vector Machine has relatively high time complexity and cannot meet the requirements of real-time processing. Based on the Pyramid algorithm, we can obtain a low-resolution representation of the images being processed. Then training and testing are applied in the low-resolution images only. Through this method, the time consumption is significantly lower than the original Support Vector Machine Procedure. The objects in farmland images are large and there are only two major types of structures in them, so the examination accuracy of the proposed method is changed little. At the same time, based on spatial structure and color distribution of the farmland image, the kernel and main parameters of Support Vector Machine are selected.