Dissertation > Excellent graduate degree dissertation topics show
Video-Based Detection of Abnormal Behavior in the Examination Room
Author: LuYong
Tutor: HeDongJian
School: Northwest University of Science and Technology
Course: Applied Computer Technology
Keywords: Abnormal behavior Video Segmentation Motion detection Time and space template Template matching
CLC: TP274
Type: Master's thesis
Year: 2010
Downloads: 130
Quote: 1
Read: Download Dissertation
Abstract
|
Electronic the invigilator system in our country has been widely used , but the examination room video surveillance still use manual observation and recording methods . Human surveillance is not only time-consuming , prone to undetected presence of subjectivity and inaccuracies defects and the judgment result . In this paper, the examination room video analysis applications demand , combined with computer vision technology , video analysis techniques and pattern recognition techniques to study effectively identify abnormal behavior based on video analysis of the examination room . This paper studies the work and conclusions are as follows : (1) intelligent monitoring application needs of the examination room , and the overall program to identify abnormal behavior of video - based exam . Design of hardware and software components of the system , as well as related settings to determine a reasonable and effective examination room video capture program . Against the background of chaos , interfere with movement detection target exist segmentation difficulties caused by partial occlusion problem , improve the segmentation accuracy analysis , comparing the advantages and disadvantages of various video segmentation method based on the use of space-time integration of the video effectively split method. ( 2 ) study using interactive video segmentation method to effectively split the training video , and to extract 3D space-time template . The interactive segmentation joined human intervention to be able to make a precise segmentation of video moving target boundary . Research based on unsupervised clustering method split test video . Calculated for the large, slow implementation proposed 8 -tree pretreatment and hierarchical clustering methods , improve segmentation efficiency . The experimental results show , the use of improved Mean Shift clustering algorithm in the case does not affect the effect of segmentation , segmentation efficiency by 2-3 orders of magnitude . (3) based on the shape of the space-time template matching method . Match in the texture - rich region prone to mismatching problem , using a combination of optical flow information supplement . Experimental results show that the improved method to reduce the error rate . ( 4 ) The choice of the Linux operating system , using the Matlab simulation system and the C programming language , combined with the Intel OpenCV design video-based exam system of intelligent detection of abnormal behavior , efficient video segmentation , detection and recognition of video moving objects .
|
Related Dissertations
- Fundus Image Segmentation Based on SVM and Template Matching,TP391.41
- Visual Servoing Approaches Based on Parallel Mechanism,TP391.41
- Multi-threaded fusion soccer video semantic analysis and event detection,TP391.41
- Technologies of Real-time Traffic Sign Segmentation and Recognition,TP391.41
- Design of Low Power Pipelined ADC for Motion-Detection CMOS Image Sensor,TN792
- The Engineering Implementation of the Algorithm for Fast Positioning in Vision by Arbitrary Angle Rotation,TP391.41
- Video Object Segmentation Based on GA in Video Surveillance,TP391.41
- Research on Vision Detection Algorithm for Tracing Printing System,TP391.41
- Application of Motion Detection Technology in Intelligent Surveillance System,TP277
- Digital rapid detection and identification of the ball,TP391.41
- Design and Implementation of Embedded Decoder System Based on Video Surveillance,TN764
- UAV-based airborne television imaging ground target tracking,TP391.41
- Animation based on partial least squares filtering research,TP391.41
- Research on Detectiong Method of Traffic Illegal Parking Based on Vision,TP391.41
- Intelligent Moving Electronic Police System,TP391.41
- Intelligent Video Analysis Technology for Elevator Cage Abnormal Behavior Detection Based on Computer Vision,TP274.4
- Design and Implementation of Hand Motion Detective System Based on Motion History Image,TP274.4
- Multiple Target Tracking with Video Segmentation and Particle Filter,TP391.41
- Integration of multiple video flame detection feature extraction studies,TP391.41
- Improvement Research and Implementation of License Plate Recognition,TP391.41
- CF1000 type Sorter image recognition system design and implementation,TP391.41
CLC: > Industrial Technology > Automation technology,computer technology > Automation technology and equipment > Automation systems > Data processing, data processing system
© 2012 www.DissertationTopic.Net Mobile
|