摘要
本文介绍了一种基于Canny算子边缘提取和Harris角点提取的三维重建方法。通过双目相机获取所研究物体的左右视图,并对图像进行平滑去噪处理,将Canny算子边缘检测和Harris角点提取结合起来进行特征提取,以提高特征点选取的准确性,然后将左右视图的特征点进行立体匹配,计算匹配点的三维坐标值,最后使用经典Delaunay三角剖分算法进行网格重建,实现了将从不同位置获取的二维图像经过计算处理得到其三维信息并进行建模的过程。
Introducting a method of three-dimensional reconstruction based on the Canny edge detection operator and the Harris comer detection .Using Bumblebee2 to get the left and right view of studied object.Then,smoothing the image,combining the Canny edge detection operator with the Harris comer detection to improve the accuracy of selected feature points,matching the feature points of the left and right view, calculating the value of matching points＇ three-dimensional coordinates .Finally, the classical Delaunay triangulation for mesh reconstruction from different locations to achieve the process ,which is a two-dimensional images obtained through the calculation by the threedimensional information and modeling.
出处
《电子测试》
2011年第2期11-16,共6页
Electronic Test
基金
山西省自然科学基金(2010011023-1)资助
关键词
双目视觉
边缘检测
特征提取
三维重建
binocular vision
edge detection
feature extraction
Three-dimensional reconstruction
