摘要
针对视频拼接实时性的需求,提出一种利用小区域融合和查表映射实现实时视频拼接的方法.该方法首先利用颜色校正算法对待拼接摄像头视频进行颜色校正;然后在初始化阶段估算出图像变换参数和小区域融合权值,并以索引表的方式保存;最后在视频拼接阶段利用查表法对视频进行实时拼接.实验结果表明,该方法在较好消除残影和拼接缝隙的同时,能将每帧的平均拼接时间降低至约0.03s,并在拼接速度和拼接效果上都较好.
According to the need of real-time video splicing, we proposed a method of real-time video splicing based on small region fusion and look-up table mapping. Firstly, we used color correction algorithm to deal with color correction of the mosaic camera video. Secondly, we estimated the image transformation parameters and the small region fusion weights at the initialization stage and stored image in the form of index tables. Finally, we used look-up table method for real-time video splicing in the video mosaic stage. The experimental results show that this method can well eliminate ghosting and splicing gap, meanwhile, the average splicing time of each frame is reduced to about 0.03 s. The method has good performance in speed and effect of video splicing.
出处
《吉林大学学报(理学版)》
CAS
CSCD
北大核心
2016年第6期1367-1372,共6页
Journal of Jilin University:Science Edition
基金
河南省科技厅科技攻关项目(批准号:112102210373)
河南省教育厅自然科学研究计划项目(批准号:2010A520003)
关键词
视频拼接
小区域融合
颜色校正
video splicing
small region fusion
color correction