Monocular vision-based robot guidance is characterized by its simple structure, flexible configuration, and ease of integration, which can effectively estimate the target pose information in the presence of cooperative markers. In this chapter, the main forms of cooperative markers for monocular vision are introduced, and the mainstream algorithms for corner, ellipse and other feature detection are explained, followed by several comparative experiments on feature detection. Moreover, the multi-view image matching problem is discussed by using SIFT algorithm and its improved versions. The pose estimation methods based on P3P and RPnP are particularly introduced, which can be further demonstrated with comparative experiments on mainstream pose solution methods.