tracker_release
所属分类：数值算法/人工智能
开发工具：matlab
文件大小：4331KB
下载次数：14
上传日期：2019-06-02 15:32:23
上 传 者：
大头猫神
说明： 基于帧图像序列的目标检测与跟踪，利用了图像分割融合的算法，有主函数可以直接运行，有数据集
(Target detection and tracking based on frame image sequence, using image segmentation and fusion algorithm, has main function to run directly, has data set)
文件列表:
cstTracker.m (8728, 2019-05-27)
data11.bmp (18486, 2019-05-27)
data12.bmp (18486, 2019-05-27)
dense_gauss_kernel.m (1080, 2012-08-01)
get_subwindow.m (1152, 2012-08-01)
load_video_info.m (2903, 2012-08-08)
run_tracker.m (3540, 2012-08-08)
show_precision.m (1245, 2012-08-08)
data\original00002.txt (18688, 2016-11-07)
data\original00003.txt (18688, 2016-11-07)
data\original00004.txt (18688, 2016-11-07)
data\original00005.txt (18688, 2016-11-07)
data\original00006.txt (18688, 2016-11-07)
data\original00007.txt (18688, 2016-11-07)
data\original00008.txt (18688, 2016-11-07)
data\original00009.txt (18688, 2016-11-07)
data\original00010.txt (18688, 2016-11-07)
data\original00011.txt (18688, 2016-11-07)
data\original00012.txt (18688, 2016-11-07)
data\original00013.txt (18688, 2016-11-07)
data\original00014.txt (18688, 2016-11-07)
data\original00015.txt (18688, 2016-11-07)
data\original00016.txt (18688, 2016-11-07)
data\original00017.txt (18688, 2016-11-07)
data\original00018.txt (18688, 2016-11-07)
data\original00019.txt (18688, 2016-11-07)
data\original00020.txt (18688, 2016-11-07)
data\original00021.txt (18688, 2016-11-07)
data\original00022.txt (18688, 2016-11-07)
data\original00023.txt (18688, 2016-11-07)
data\original00024.txt (18688, 2016-11-07)
data\original00025.txt (18688, 2016-11-07)
data\original00026.txt (18688, 2016-11-07)
data\original00027.txt (18688, 2016-11-07)
data\original00028.txt (18688, 2016-11-07)
data\original00029.txt (18688, 2016-11-07)
data\original00030.txt (18688, 2016-11-07)
data\original00031.txt (18688, 2016-11-07)
data\original00032.txt (18688, 2016-11-07)
... ...
This MATLAB code implements a GUI for the tracking algorithm described in:
Joo F. Henriques, Rui Caseiro, Pedro Martins, and Jorge Batista,
"Exploiting the Circulant Structure of Tracking-by-detection with Kernels,"
ECCV, 2012.
It is free for research use. If you find it useful, please acknowledge the
paper above with a reference.
The main script for tracking is "run_tracker.m". It should work with any
recent version of MATLAB (eg., 2009a and above). It only requires the
Signal Processing toolbox, which is fairly standard.
You probably want to change the value of "base_path" at the top of the main
script to your videos' location. Each video is a folder with appropriate
files, in MILTrack's format. They can be downloaded from:
http://vision.ucsd.edu/~bbabenko/project_miltrack.shtml
Alternatively, you can use any folder with PNG images as a video. It must
contain a text file in MILTrack's ground truth format, with bounding box
coordinates for at least the first frame, to initialize the tracker.
Joo F. Henriques, 2012
http://www.isr.uc.pt/~henriques/
近期下载者：
相关文件：
收藏者：