边界重叠图像的网格拼接算法

毕业设计（论文）说明书

题目：边界重叠图像的网格拼接算法

系名计算机科学与技术系

专业软件工程

学号 6007203156

姓名徐御臣

指导教师殷妍

2011年6月3日

图像拼接技术是将一组相互间重叠部分的图像序列进行空间匹配对准，经重采样合成后形成一幅包含各图像序列信息的宽视角场景的、完整的、高清晰的新图像的技术。图像拼接在摄影测量学、计算机视觉、遥感图像处理、医学图像分析、计算机图形学等领域有着广泛的应用价值。一般来说，图像拼接的过程由图像获取，图像配准，图像合成三步骤组成，其中图像配准是整个图像拼接的基础。针对边界部分有重叠的图像，提出了一种基于网格匹配的快速对准算法，并通过平滑因子对图像实现了无缝拼接。提出了一种适宜于生成全方位全景图像的拼接和平滑算法，给出了一种消除拼接积累误差的方法，进而首次在立方体表面拼接成功了全方位全景图像，并由此生成了球面上的全方位全景图像。

关键词：图像拼接；图像配准；图像融合；全景图

Image mosaic is a technology that carries on the spatial matching to a series of image which are overlapped with each other, and finally builds a seamless and high quality image which has high resolution and big eyeshot. Image mosaic has widely applications in the fields of photogrammetry, computer vision, remote sensing image processing, medical image analysis, computer graphic and so on. In general, the process of image mosaic by the image acquisition, image registration, image synthesis of three steps, one of image registration are the basis of the entire image mosaic. In this paper, we present a fast stitching algorithm for overlapping images based on grid matching, which makes images matching correctly, stitching images seamless and smooth. Both a mosaic method and a smoothing method that are adaptive to be used to create omni-directional images are proposed. A method by which the mosaic accumulative error can be eliminated is also proposed. As the first time, an

omni-directional image is created on cube face, through which an omni-directional image on a sphere face is created.

Keywords: image mosaic; image registration; image fusion; panorama