----------------------
Citation Details
----------------------
Please cite the following journal article when using this source code:
V. Reddy, C. Sanderson, B.C. Lovell.
Improved Foreground Detection via Block-based Classifier Cascade with Probabilistic Decision Integration.
IEEE Transactions on Circuits and Systems for Video Technology, Vol. 23, No. 1, pp. 83-93, 2013.
DOI: 10.1109/TCSVT.2012.2203199
You can obtain a copy of this article via:
http://dx.doi.org/10.1109/TCSVT.2012.2203199
----------------------
License
----------------------
The source code is provided without any warranty of fitness for any purpose.
You can redistribute it and/or modify it under the terms of the
GNU General Public License (GPL) as published by the Free Software Foundation,
either version 3 of the License or (at your option) any later version.
A copy of the GPL license is provided in the "GPL.txt" file.
----------------------
Instructions and Notes
----------------------
To run the code the following libraries must be installed:
1. OpenCV 2.4 (later versions should also work)
2. Armadillo 3.920 - http://arma.sourceforge.net
Under Linux, to compile the code use the following command:
g++ -o ForegroundSegmentation main.cpp input_preprocessor.cpp -O2 -fopenmp -I/usr/include/opencv -I/usr/local/include/opencv -L/usr/lib64 -L/usr/local/lib -larmadillo -lopencv_core -lopencv_highgui -lopencv_imgproc
You may need to adapt the paths for libraries and includes to suit your environment.
The above command line has been tested on Fedora 19, using Armadillo 3.920.2 and OpenCV 2.4.6
After successful compilation, to execute the code, run the following command:
./ForegroundSegmentation <set input path sequence> <sequence name>
For example:
./ForegroundSegmentation /home/Project/datasets/UCSD/seq1/ seq1
Points to note:
1.
Supported input formats: png, jpeg, bmp and tif.
2.
Internally, the code sorts the input image files of a given folder in ascending order.
Hence, the file names must contain a constant number of digits in their suffixes
(eg. test_0001, test_0002, test_0100, test_1000,...).
3.
Initially, the algorithm uses first 200 frames to build a model of the background
before producing foreground mask for each frame.
4.
To save the masks, WRITEMASK must be defined in main.hpp (by default, this is defined).
An output folder is automatically created to store all the generated foreground masks.
The output masks are stored as png images.
5.
The code is currently not optimised for speed.
没有合适的资源?快使用搜索试试~ 我知道了~
相机抖动下前景提取代码
共18个文件
hpp:14个
txt:2个
cpp:2个
需积分: 12 12 下载量 4 浏览量
2017-09-20
14:48:49
上传
评论
收藏 43KB ZIP 举报
温馨提示
智能化视频监控系统的研究逐渐成为近年来计算机视觉领域研究的一个重要方向,而视频中的前景提取是智能监控技术中的一个至关重要的课题,也是研究者最感兴趣的问题之一。视频中的前景提取涉及到诸如人工智能、图像处理、模式识别、视频分析等多个学科的交叉领域,因此是一个极具挑战性的系统工程问题。对于视频中运动前景部分提取,本代码采用了一种基于多通道高斯混合模型的动态背景模型。通过对背景模型的改进,提高了前景提取的准确性,并使其能够满足实时检测的要求。针对提取前景出现细小空洞以及不连贯问题,本代码采取了图像形态学的处理,增强了前景提取的可靠性。
资源推荐
资源详情
资源评论
收起资源包目录
foreground_detection_code.zip (18个子文件)
foreground_detection_code
foreground_detection_code
README.txt 3KB
GPL.txt 35KB
code
input_preprocessor.hpp 2KB
CascadedBgsParams_proto.hpp 1KB
mog_diag_em_ml_parallel_meat.hpp 10KB
CascadedBgs_meat.hpp 26KB
mog_diag.hpp 821B
main.hpp 883B
mog_diag_kmeans_parallel_proto.hpp 2KB
CascadedBgs_proto.hpp 3KB
main.cpp 6KB
input_preprocessor.cpp 6KB
mog_diag_base_proto.hpp 3KB
inc.hpp 840B
CascadedBgsParams_meat.hpp 1KB
mog_diag_em_ml_parallel_proto.hpp 2KB
mog_diag_base_meat.hpp 15KB
mog_diag_kmeans_parallel_meat.hpp 9KB
共 18 条
- 1
资源评论
SUNYALILIANG
- 粉丝: 0
- 资源: 1
上传资源 快速赚钱
- 我的内容管理 展开
- 我的资源 快来上传第一个资源
- 我的收益 登录查看自己的收益
- 我的积分 登录查看自己的积分
- 我的C币 登录后查看C币余额
- 我的收藏
- 我的下载
- 下载帮助
最新资源
- 杨辉三角(Pascal's Triangle)是一个在数学中非常著名的数列,它以一种特定的模式排列数字,内容以5种代码实例实现
- yolov5数据集:CT图像肾脏、结石检测(2类别,包含训练集、验证集)
- 电力102规约请求报文实录
- RTL8723 linux+安卓驱动 2020版本
- ISO 26262 道路车辆-功能安全-系统级产品开发
- 数据挖掘期末题 选择填空简答
- EKF在CTRV模型上的建模推导
- 用于:https://laoshifu.blog.csdn.net/article/details/138191428 多层创建
- 电力102规约(电能累计量的传输规约)培训
- 「常用」E9外部流程接口
资源上传下载、课程学习等过程中有任何疑问或建议,欢迎提出宝贵意见哦~我们会及时处理!
点击此处反馈
安全验证
文档复制为VIP权益,开通VIP直接复制
信息提交成功