基于支持向量机与细节层次的三维地形识别与检索

3D Terrain Recognition and Retrieval Based on Support Vector Machine and Level of Detail

提出对相似 3D物体识别与检索的算法 该算法首先使用细节层次模型对 3D物体进行三角面片数量的约减 ,然后提取 3D物体的特征 由于所提取的特征维数很大 ,因此独立成分分析被用来进行 3D特征约减 基于约减后的特征 ,使用支持向量机进行识别与检索 将该算法用于 3D丘陵与山地的地形识别中 。

D features are first filtered by reducing the meshes of 3D object through level of detail. Since the dimension of extracted 3D features are very huge, the independent component analysis (ICA) is used to exaggerate the main and most obvious features, then the recognition of similar 3D objects is realized by support vector machine (SVM). The proposed algorithm works well when it is used to recognize and retrieve 3D terrain.etri