摘要
针对传统的混合高斯模型不能很好地处理突变的情况,提出了一种新的运动目标检测算法。该算法在时间域上对混合高斯模型的更新机制进行了改进,并对模型加入了帧间处理,提取出初步的运动目标;在空间域上,通过Canny边缘检测算子获得初步的运动目标边缘轮廓,利用图像金字塔的多分辨特性排除背景噪声,经过一定运算再次得到运动目标。将两次得到的运动目标作"与运算",提取出最终的运动目标。实验结果表明,本算法可以较好地处理突变情况,提取的运动目标图像完整、轮廓清晰。
Aiming at the fact that the traditional Gaussian mixture model can not deal with the mutations effectively, this paper proposes an improved moving object detection algorithm. The model＇s updating mechanism is improved on the time domain, while joining frame disposal to the model, and obtain the preliminary moving object. Canny edge detection operator is used to get the edge figure on the space domain, and extract the preliminary outline of the moving object. Using the multi-resolution characteristics of the image-pyramid to eliminate the noise in background region, perform further operation, and extract the preliminary moving object again. Obtain the result by logic and processing for these two objects. Experiments show that this algorithm can effectively handle the mutation, and improve motion detection precision.
出处
《微型机与应用》
2011年第23期43-45,50,共4页
Microcomputer & Its Applications
基金
国家自然科学基金(60778007)
关键词
边缘检测
混合高斯模型
运动目标检测
背景减除
edge detection
Guassian mixture model
moving object detection
background subtraction
