In various fields including 3D reconstruction , multidimensional data fusion and information visualization , in the stitching process of real-time video data , the splicing gap problem often appears .An improved weighted fusion algorithm for real-time video stitching is presented .In this algorithm ,the problem of video stitching is transformed into the corresponding frame images stitching . The image stitching algorithm is based on SURF (Speed Up Robust Features) . Firstly , SURF is used for extracting and matching feature points . Then , the geometric transform matrix of image based on RANSAC(Random Sample Consensus) algorithm and LM (Levenberg-Marquard) algorithm is calculated .Interpolation operation is performed for stitching the images with the transform matrix .Finally ,the image content of the stitching region is weighted to complete the stitched image fusion .In the fusion process ,the merging line side based weighted fusion and image content based weighted fusion methods are proposed . Experimental results show that the proposed fusion algorithm has the advantages of fast processing speed and good effect ,and it is suitable for the scene video stitching and real-time data presentation with requirement of the stitching speed and the effect .%在三维重建、多维数据融合和信息可视化分析等领域中,针对实时视频图像数据拼接融合的过程中常出现的拼接缝隙问题,提出了一种改进的加权融合实时图像拼接方法.该方法将视频拼接问题转化为视频对应帧图像的融合问题,通过利用S U RF算法对特征数据点进行提取与匹配,对图像使用RANSAC算法和LM算法求其几何变换矩阵,再根据变换矩阵进行图像插值拼接,对拼接线区域图像内容进行加权处理完成视频图像融合.在进行图像融合处理时,提出了基于拼接线一侧加权融合和基于原图内容的加权融合算法.实验结果表明:该方法实现的拼接融合速度快、效果较好,适用于对速度和拼接效果均有要求的场景视频拼接和实时数据呈现.
展开▼
