摘要
针对低照度、雾霾情况下的视频信息有限,实时性要求高等问题,提出基于DSP的嵌入式实时视频配准和网络编码传输方案。该系统以DM642开发板为硬件平台,使用TI官方DSP/BIOS实时嵌入式系统。通过红外与可见光摄像头实时采集图像数据,采用基于SURF算法的图像配准,计算Hessian矩阵确定对应匹配点对,通过矩阵插值获得最后配准图像。该方法速度快,满足实时配准需要,在配准后实现了H.264算法的实时视频的编码压缩、解码传输以及通过网络进行发送。
Aiming at the problem of limited video information and high real-time requirement under low illumination and haze conditions,a realtime video registration and network coding transmission scheme based on DSP is proposed.The system uses DM642development board as the hardware platform,and uses TI official DSP/BIOS real-time embedded system.The image data are collected in real time by infrared and visible camera.Image registration based on SURF algorithm is used to calculate Hessian matrix to determine the corresponding matching points.Finally,the final registration image is obtained through matrix interpolation.The method is fast and real-time registration.After registration,the real-time video coding,compression,decoding and transmission of H.264algorithm and the transmission through the network are realized.
作者
任思阳
Ren Siyang(Electronic and Information Engineering Institute,Changchun University of Sciences and Technology,Changchun Jilin 130022)
出处
《数字技术与应用》
2017年第11期131-132,共2页
Digital Technology & Application