搜索资源列表
gmm
- 利用混合高斯模型训练视频,获得背景图像,并将背景保存。-training video with GMM model ,then get the background,and store the picture in your computer.
intro
- IBM官方COGNOS 培训视频1(介绍)-IBM official COGNOS Training Video 1 (Introduction)
manager
- IBM官方COGNOS 培训视频2(管理)-IBM official COGNOS Training Video 2 (Management)
modeller
- IBM官方COGNOS 培训视频3(模式)-IBM official COGNOS Training Video 3 (mode)
analyst
- IBM官方COGNOS 培训视频4(分析)-IBM official COGNOS Training Video 4 (analysis)
biprof
- IBM官方COGNOS 培训视频5(biprof)-IBM official COGNOS Training Video 5 (biprof)
gnebehay-OpenTLD-v1.3.0-10-ge50d202
- 纯C++版本的OpenTLD-v1.3.0,效果不错。-This is a C++ implementation of OpenTLD that was originally published in MATLAB by Zdenek Kalal. OpenTLD is used for tracking objects in video streams. What makes this algorithm outstanding is that it does not make use of
GMM
- 针对摄像机固定下的复杂背景环境,对采集到的视频图像的图像数据用混合高斯背景建模方法实现前景/背景分割,实现运动目标检测和跟踪。在进行前景检测前,先对背景进行训练,对图像中每个背景采用一个混合高斯模型进行模拟,每个背景的混合高斯的个数可以自适应。然后在测试阶段,对新来的像素进行GMM匹配,如果该像素值能够匹配其中一个高斯,则认为是背景,否则认为是前景。由于整个过程GMM模型在不断更新学习中,所以对动态背景有一定的鲁棒性。最后通过对一个有树枝摇摆的动态背景进行前景检测,取得了较好的效果。-For c