基于FPGA的全景视频图像拼接的设计与实现  

The design and implementation for the splicing of the panoramic video images based on FPGA

作　　者：李勇[1] 王磊[1] 钱罕林[1] 

机构地区：[1]新疆轻工职业技术学院信息与软件分院,新疆乌鲁木齐830021

出　　处：《电子设计工程》2018年第2期80-83,共4页Electronic Design Engineering

摘　　要：针对多端全景摄像机在视频图像拼接过程中的耗时长的问题,探索采用一种设计方法,通过从算法和系统优化设计入手,采用简化的SIFT算法,并且将FPGA平台与此算法计算分离,转而由ARM来承担图像配准参数的计算,由FPGA平台依据参数结果进行全景视频图像的拼接。本文涉及了应用FPGA系统设计实现对具体参数的接收、计算和对视频图像的缩放、平移以及最终的仿真调试等。实验表明:全景视频图像经过这种方法拼接后能实时显示在屏幕上且效果良好。To overcome the time-consuming problem encountered in multi-camera panoramic imaging system,starting from algorithm and system optimization,using the simplified SIFT algorithm,separating the FPGA platform and the calculation,the image registration parameters are calculated by ARM,according to the result of the parameter,the FPGA system is used to realize the panoramic video mosaic.The design is involved in the video image scaling,translation and reception of the parameters,the calculation module design and system simulation debugging,etc. The experimental results show that the panoramic video images can be displayed on the screen in real time and the effect is good.

关 键 词：FPGA 全景视频 图像拼接 SIFT算法 

分 类 号：TN911.73[电子电信—通信与信息系统]