The 3D Model Retrieval Based on Local Features

When using global features for 3D model retrieval, we ignore the local similarity of 3D model. To solve this problem, we propose a 3D model retrieval method based on local features. First, calculate the normal of each vertex, get the normal projection of each vertex and identify salient points according to the SURF feature of the projection. Extract the local feature around the salient points obtained, and use the local feature to update the global similarity. The experimental results show that this method can effectively improve the 3D model retrieval precision.