摘要
近年来,车辆的主动安全研究已经引起广泛的关注。针对汽车主动安全中行人检测的需求,基于彩色差值模型算法,通过Gauss算法分割图像和kalman滤波预测目标在下一时间可能出现的活动区域,提出了一种基于运动趋势估计的行人检测算法。经仿真表明,该汽车动态目标检测算法可实现行人运动趋势的估算,具有较好的实时性、可靠性。
In recent years,the active safety research of vehicle has aroused widespread attention.In view of the pedestrian detec tion in automotive active safely requirements and based on color difference model algorithm,the article puts on the movement trend estimation through the Gauss algorithm segmentation image and Kalman filtering prediction target areas of activity that may apper in the next time.The simulation results show that the venicle dynamic target detection algorithm can realize the pedestrian movement trend estimation.It has good read-time performance and reliability.
出处
《电脑知识与技术》
2013年第1X期571-574,共4页
Computer Knowledge and Technology
基金
高职院校信息技术类课程实践教学体系改革与实践(2012jyxm645)
关键词
行人检测
图像差分
图像分割
KALMAN滤波
pedestrian detection
image difference
image segmentation
kalman filtering
