基于未校准照片构造全景图的新方法

提出一种基于未校准的系列照片构造全景图纳新方法，该组照片可采用普通三脚架/数码相机在固定地点拍摄。采用普通三角架通过旋转角度拍摄的照片难以保证轴心一致性，可能出现图像扭曲、交叠和倾斜，照片之间可能有一定色差，因此，在图像的拼接和建立全景图方面难度大。文中采用特征匹配技术实现照片的拼接，通过照片的定位拼接、相邻照片颜色亮度的协调实现全景图的构建。该算法所生成的全景图效果好，计算速度快，巳得到实际应用。它可广泛应用于Web环境下的基于图像的虚拟场景构造等方面。

A New Method for Building Panoramas from Uncalibrated Photographs

In this paper a new method is presented to build full view cylindrical panorama from a group uncalibrated photographs. The photos can be taken naturally with an ordinary digital camera supported by a tripod at an (approximately) fixed location. Such photos usually have large perspective distortion, small overlap, brightness difference, small concentric errors and camera rotations. These characteristics make both image alignment and panorama building more difficult than using photos taken by cameras calibrated by special equipment. Simple registration techniques cannot be applied directly to align such images. The novel image registration approach based on feature matching scheme is proposed in this paper. The panorama can be build by two steps: image registration and brightness adjustment. Experiments show that our methods yields good results and has higher computing speed. The method has been practically used in digital museum project.