抗畸变高效的视频拼接算法

摘要

A novel anti-distortion and efficient video stitching algorithm is proposed to solve video splicing edge distortion and fusion efficiency.The establishment of nonlinear anti-distortion model (NAM)improves the smoothness of the transition edge by correcting the non-linear distortion of the video image.In order to solve the contradiction between accuracy and effi-ciency,speeded up robust features (SURF)algorithm is used.And two-dimensional mark model (2DMM)is presented to accelerate the speed of image fusion by reducing redundant computation.Compared with the traditional method,it promotes the video splicing quality and fusion efficiency.%针对视频拼接存在的边缘失真和融合效率问题，提出一种新的抗畸变高效的实时视频拼接算法。通过建立非线性抗畸变模型，对视频图像存在的非线性畸变进行校正，改善图像边缘过渡的平滑性；采用 SURF配准算法兼顾配准的精度和效率；使用二维标记模板降低融合过程的冗余计算，进行快速图像融合。该算法提升了传统方法的视频拼接质量与融合效率。