摘要随着现代社会监控难度的增加,越来越多的领域对传统意义上的视频监控提出了更高的要求,智能视觉监控技术旨在赋予监控系统观察分析场景内容的能力,实现监控的自动化和智能化。其智能化分析过程由运动目标检测、分类、跟踪和视频内容分析等环节组成,从视频序列中检测运动目标是一个基础而又关键的任务。
红外与可见光视频传感器输出信息具有互补性,协同利用优势突出,使得全天时全天候鲁棒智能监控成为可能。本课题研究同时利用可见光和红外视频信息,通过视频处理算法实现运动目标的检测。
论文主要工作包括:1.设计了一种基于背景模型的融合检测方法。首先结合像素在时空域上的分布信息构造时空模型,给出了模型更新方法和前景检测方法,然后利用该模型对可见光和红外图像序列分别进行建模和前景检测,给出了一种基于置信度的目标融合检测方法,利用双传感器信息提高检测精度和可靠性,仿真实验验证了其有效性。2.设计实现了基于Matlab的融合检测软件,通过将GUI界面软件和检测算法结合,得到了预期结果。5765
关键词 智能视觉监控 目标检测 多模式均值时空模型 融合检测
毕业设计说明书(论文)外文摘要
Title Visible -infrared video target fusion detection software design and implementation
Abstract
With the increase of modern social control of the difficulty, more and more areas of video surveillance in the traditional sense of the higher requirements for intelligent visual surveillance technology designed to give the monitoring system to observe the analysis of scene content, automate monitoring andintelligent. Its intelligent analysis of the process of moving target detection, classification, tracking and video content analysis and other aspects of composition, the detection of moving targets from a video sequence is a basic and crucial task.
Complementary infrared and visible video sensor output, collaborative use of the advantages highlighted, making the all-day all-weather Robust intelligent monitoring become possible. This research using visible and infrared video, and video processing algorithms to detect the moving object.
Study include: 1. Designed a fusion detection based on background model. First combine the spatial and temporal model of the distribution of pixels in the time-space domain construct, model update method is given and the prospects of detection methods, then using this model for modeling and foreground detection of visible and infrared image sequences, respectively, is given based on confidence target fusion detection using dual-sensor information to improve the detection accuracy and reliability. Simulation results show its effectiveness. (2) Design and Implementation of the integration testing software based on Matlab GUI interface software and detection algorithm combined with the expected results.
Keywords Intelligent vision surveillance target detection multi-mode mean space-time model integration testing
目 次
1 绪论 1
1.1 智能视频监控 1
1.2 智能监控技术研究内容 1
1.3 运动目标检测方法及现状 3
1.4 论文的主要内容及工作安排 5
2 MATLAB GUI界面编程简介 6
2.1 GUI介绍 6
2.2 GUI设计规范 7
2.3 GUI设计实现 9
3 多模式均值时空模型的目标融合检测方法 11
3.1 引言 11
3.2 多模式均值时空模型 12
- 上一篇:基于PLC的地铁站自动排水控制系统设计+梯形图
- 下一篇:Thevenin动力电池SOC估计研究+文献综述
-
-
-
-
-
-
-
g-C3N4光催化剂的制备和光催化性能研究
上市公司股权结构对经营绩效的影响研究
中国传统元素在游戏角色...
高警觉工作人群的元情绪...
浅析中国古代宗法制度
江苏省某高中学生体质现状的调查研究
现代简约美式风格在室内家装中的运用
巴金《激流三部曲》高觉新的悲剧命运
NFC协议物理层的软件实现+文献综述
C++最短路径算法研究和程序设计