A new visual navigation method is proposed in the paper, which take advantage of natural landmarks and image local features. Images of natural landmarks are first collected and organized as database. SURF features of natural landmarks are also extracted and saved. In practice, real-time images are captured by camera, and their SURF features are also extracted. For each real-time image, its SURF features are used to match with corresponding ones of images of natural landmarks. According to match rules, it can be decided that which landmark the real-time image belongs to. Experimental results demonstrate that this method has high accuracy and strong robustness in complex environment.
|