融合码本模型和边缘检测的目标检测算法  

Target detection algorithm based on codebook model and edge detection

作　　者：许锦婷 陈仁文[1] 余小庆[1] Xu Jinting;Chen Renwen;Yu Xiaoqing(Mechanical structural laboratory, Nanjing University of Aeronautics and Astronautics, Nanjing 210016,China)

机构地区：[1]南京航空航天大学机械力学结构实验室

出　　处：《国外电子测量技术》2018年第5期141-145,共5页Foreign Electronic Measurement Technology

摘　　要：提出了一种基于码本模型和边缘检测的运动目标检测算法。首先对基础码本模型中的码字部分进行改进,利用像素点中k个最大、最小Y分量的平均值来代替最大、最小Y分量值;同时融入自适应卡尔曼滤波的思想,利用前后两帧图像的亮度信息对当前像素值进行预测及修正;最后将改进后的码本模型算法与边缘检测相结合,利用码本模型检测到的初始运动目标与边缘检测检测到的运动目标相加得到最终的运动目标。实验结果表明,该改进算法在前景提取率上相比GMM算法提高了24%,且实时处理方面也能达到57帧/s,可以满足实际应用的实时性要求。This paper presents a moving object detection algorithm based on codebook model and edge detection.Firstly,the part of the code in the basic codebook model is improved,and the average value of the klargest and smallest Ycomponents in the pixel is used to replace the maximum and minimumYcomponent values.At the same time,the idea of adaptive Kalman filter is used,the brightness information of the frame image is used to predict and correct the current pixel value;finally,the improved codebook model algorithm is combined with the edge detection,and the initial moving target detected by the codebook model is added with the moving target detected by the edge detection The ultimate sporting goal.The experimental results show that the improved algorithm in this paper improves the foreground extraction rate by24% compared with the GMM algorithm and can achieve 57 frames/second in real-time processing,which can meet the real-time requirements of practical applications.

关 键 词：码本模型 自适应卡尔曼滤波 边缘检测 

分 类 号：TN911.73[电子电信—通信与信息系统]