摘要
本文提出利用可见光传感器、红外传感器设备的多源融合监测方案,利用光学测量设备捕捉图像帧序列的颜色、形状、纹理、边缘细节等信息,采用扩展卡尔曼滤波(Extended Kalman Filter,EKF)算法对目标对象的像素灰度值作出预处理、去噪,基于Otsu算法的图像动态目标形状特征,将可见光与红外传感器的跟踪目标样本位置作出比对,得到相距最近的位置作为跟踪结果,由此提升复杂光照或环境条件下目标多源融合跟踪的识别准确率.
Abstract
This paperPropose a multi-source fusion monitoring scheme using visible light sensor,infrared sensor equipment,using optical measurement equipment to capture the color,shape,texture,edge details and other information of the image frame sequence,using the Extended Kalman Filter(EKF)algorithm to make preprocessing,denoising of the pixel gray value of the target object,based on Otsu algorithm of Based on the Otsu algorithm,the dynamic target shape characteristics of the image,the visible light and infrared sensor tracking target sample location to compare,to get the closest position as the tracking results,thus improving the recognition accuracy of multi-source fusion tracking of targets under complex lighting or environmental conditions.
国家科技期刊平台
NSTL
万方数据