Object Recognition with Corner-Based Feature

Object recognition with changes in 3D viewpoint is a hot spot currently in the domain of image pattern recognition. An algorithm for pattern matching based on Hausdorff distance is proposed in this paper. The principal components transformation is introduced to this algorithm and makes it stable, distinct and almost real time. Meanwhile, a method for extracting corner feature from images is proposed. The extracted feature is invariant to translation, rotation, scale change and is shown robust to addition of noise. Further, a system to recognize the object with changes in 3D viewpoint is presented using this feature and BP network. The results of experiments for comparison demonstrate that the proposed method is more effective than the other three ones.