基于蚁群搜索的三维CAD模型相似性计算

发布时间：2018-07-13 19:58

【摘要】：模型相似性计算是三维CAD模型检索中的关键技术.为了更准确地区分模型之间的差异,提出了一种基于蚁群搜索的模型相似性计算方法.首先,根据边数的差异度量源模型面与目标模型面之间的形状相似程度,并引入面邻接对应关系矩阵衡量两个模型之间的结构相似性;其次,使用蚁群算法搜索源模型与目标模型之间的最优面匹配序列,以最优面匹配序列为基础来计算两个模型之间的相似性;最后,使用贪心算法和本文所提出的方法分别计算源模型与目标模型之间的相似性,并进行对比实验.实验结果表明:在计算关键模型的相似性时,本文所提出方法的计算结果比贪心算法提高了8.33%;与贪心算法相比,本文方法能够有效区分实验中的10个模型.
[Abstract]:Model similarity calculation is a key technology in 3D CAD model retrieval. In order to distinguish the difference between models more accurately, a model similarity calculation method based on ant colony search is proposed. Firstly, according to the similarity degree of shape between the model surface and the target model surface, the similarity of structure between the two models is measured by the adjoining correspondence matrix between the two models. Ant colony algorithm is used to search the optimal surface matching sequence between the source model and the target model, and the similarity between the two models is calculated based on the optimal surface matching sequence. The similarity between the source model and the target model is calculated by greedy algorithm and the method proposed in this paper. The experimental results show that the proposed method is 8.33 higher than the greedy algorithm in calculating the similarity of the key models, and compared with the greedy algorithm, the proposed method can effectively distinguish 10 models in the experiment.
【作者单位】：哈尔滨理工大学计算机科学与技术学院;哈尔滨理工大学软件学院;哈尔滨理工大学测控技术与仪器黑龙江省高校重点实验室;
【基金】：国家自然科学基金资助项目(61502124,60903082) 中国博士后科学基金资助项目(2014M560249) 黑龙江省自然科学基金资助项目(F2015041,F201420)
【分类号】：TP391.72

【相似文献】