节点文献
单目多角度三维重建技术研究
【摘要】 本文使用多角度图片进行了三维重建技术研究,首先在二维图片中获取特征点,提出ORB-SIFT特征提取优化算法,然后基于最小生成树设计多角度匹配算法提高图片之间的立体点匹配效率,并结合三维立体视觉算法进行三维重建,实验证明与传统算法相比,本文算法时间耗费少,重建的三维模型精度提高。
【Abstract】 Multi-angle picture technology to reconstruct the three dimensional model is researched in this paper. First of all,feature points are gotten in the two-dimensional image, the ORB-SIFT feature extraction algorithm is put forward, and then multi-angle matching algorithm is improved to match stereo matching points between images effectively based on the minimum spanning tree. And 3 D algorithm is combined to reconstruct. The experiment proves that this algorithm is less time consuming,accuracy of 3 D model is improved compared with the traditional algorithm.
【关键词】 单目多角度图片;
三维重建;
立体视觉;
【Key words】 Monocular multi-angle images; Three-dimensional reconstruction; Stereo vision;
【Key words】 Monocular multi-angle images; Three-dimensional reconstruction; Stereo vision;
【基金】 齐齐哈尔市科学技术计划项目(工业攻关)(项目编号:GYGG-201613);黑龙江省教育厅基本业务专项理工面上项目(项目编号:135209243)
- 【文献出处】 科学技术创新 ,Scientific and Technological Innovation , 编辑部邮箱 ,2018年16期
- 【分类号】TP391.41
- 【被引频次】2
- 【下载频次】253