CAJ | 학술논문

在静止、复杂背景的运动目标检测算法中，减背景技术是高效的方法。大多数背景建模和背景更新运算量大，本文将目标的边缘特征融入到减背景算法中，通过对离散的目标边缘梯度进行网格密度连通域判断来提取目标。实验结果表明，该算法可以有效地减小光线变化、噪声等影响，程序优化后系统运算速度提高，能够实时地对运动目标进行检测。
재정지、복잡배경적운동목표검측산법중，감배경기술시고효적방법。대다수배경건모화배경경신운산량대，본문장목표적변연특정융입도감배경산법중，통과대리산적목표변연제도진행망격밀도련통역판단래제취목표。실험결과표명，해산법가이유효지감소광선변화、조성등영향，정서우화후계통운산속도제고，능구실시지대운동목표진행검측。
Background subtraction arithmetic is an efficient moving objects detection method based on still and complicated background. As for most of the background models＇ need of update using plenty of complicated calculation, objects are detected by judging gridding density of the objects＇ edge information. The experimental results show that the algorithm can overcome the affect of noise and light change. After program optimization, the system can detect the moving objects in real-time.