Dissertation > Excellent graduate degree dissertation topics show

Detection of Intentional Camera Movement Using Histogram and Gaussian Mixture Models

Author: XuZuo
Tutor: HuFuQiao
School: Shanghai Jiaotong University
Course: Pattern Recognition and Intelligent Systems
Keywords: intelligence surveillance camera tamper detection histogram Gaussian Mixture Modeling tracking algorithm real time robustness
CLC: TP391.41
Type: Master's thesis
Year: 2009
Downloads: 51
Quote: 3
Read: Download Dissertation

Abstract


In recent years, video surveillance system has gained importance in security and military applications. Due to the wide use of video surveillance systems, the amount of data that has to be monitored and interpreted has increased enormously. The pure mass of information that has to be handled by people has overgrown their capabilities. It’s therefore crucial to support human operators with automatic surveillance system which notify their supervisors in case of an incident potentially relevant to security. In this paper, we present an intelligence surveillance system which can automatically trigger the alarm when the monitoring camera is intentionally moved or tampered.We define the intentional camera tamper includes vision occlusion, uniformly black/white frame, over/under exposure, zoom and camera movement. The definition in image is that some properties of image such as color, texture and edge have changed and last for a period of time. For an intelligence surveillance system, the detection of intentional camera movement should address three problems: first, the method must work online in real time; second, it must detect the abnormal situation with high precision and low rate of false alarm and missing report; third, the method must have great robustness.This paper made a deep research on the detection of intentional camera movement for intelligence surveillance system. On basis of abundant inland and overseas reference papers, we made a comparison of the advantages and disadvantages of all these detection algorithms. After the research and comparison, we present the detection logic. Use this logic, we can solve the problem of making a distinction between abnormal situation and normal situation. The paper pick up the background subtraction at first, then measure the image dissimilarity based on the histogram. By building long-term pool and short-term pool to storage frames, this method is weak in large calculation and can not detect all kinds of abnormal situation. In order to improve the accuracy of the detection, we choose the Gaussian mixture modeling to distinct the background and foreground. We discuss about the parameter initialization, the parameter updating, the selection of scene background and the detection of moving foreground in Gaussian mixture modeling. Considering the robustness of the surveillance system besides the real-time and accuracy, tracking algorithm is also used in this paper. The experiments show that the detection algorithm based on the histogram and Gaussian mixture models we present here not only can make the system work online in real time, but also has great robustness to distinct the background and foreground. So, it can be used in the intelligence surveillance system which has a high demand of accuracy and time-consuming.

Related Dissertations

  1. Satellite Attitude Determination Based on Gyro and Star Sensor,V448.2
  2. Research of High Speed Image Pre-processing System Based on FPGA,TP391.41
  3. Research and Implementation on Content-Based Clothing Image Retrieval,TP391.41
  4. FPGA/DSP Image Co-processor Technology and Ethernet Data Transmission,TP391.41
  5. Development of High Speed Data Acquisation and Playback Module Based on PCI Express Bus,TP274.2
  6. Cloning and Expression of CHS and CHI Genes and Their Regulation on the Accumulation of Flavonoids in ’Cara Cara’ Navel Orange (Citrus Sinensis Osbeck) and ’Guoqing NO.4’ Satsuma Mandarin (Citrus Unshiu Marcow),S666.4
  7. Based on statistical methods of magnetic resonance imaging of the human brain image segmentation and three-dimensional analysis of the data,R445.2
  8. The Detection of the Virulence-related Gene from Vibrio Alginolyticus and the Study on Preservation Methods,S943
  9. Control of Verticillium Wilt Disease of Cotton Plants with the Application of a Bio-Organic Fertilizer and Its Microbiologecal Mechanism in Rhizosphere,S144.1
  10. Comparative Study on the Host Choice Mechanism of Helicoverpa Armigera (Hübner) and H.assulta (Guenée),S435.622.3
  11. Identifies Differential Gene Expression Related to Gravitropism-Horizontal Gravitropism in Creeping Shoots of Ground-Cover Chrysanthemum,S682.11
  12. An Adaptive Processing Node Selection Based Spatio-Temporal Query Alogrithm,TN929.5
  13. Cloning and Characterization of Five Zinc-finger Proteins (ZFPs) Related with Stresses in Cotton (Gossypium Hirsutum L.),S562
  14. Identification and Robustness Analysis of Nonlinear Hybrid Dynamical System Concerning Glycerol Transport Mechanism,TQ223.162
  15. The Study of Right Lateral Position under Real-time Monitor to Apply Capsule Endoscopy,R574
  16. Attribute Reduction of Interval-valued Information System Based on Fuzzy Discernibility Matrix,O159
  17. Isolation of Chlorobenzene-Degrading Strain and Its Application in Bioaugmenting Chlorobenzene-Treating BTF,X703
  18. The Study on the System of the Staff Positioning,TP273.5
  19. Research into Visual Orientation in the Weeding Robot Based on COCM,TP391.41
  20. Study on Network Technology of Wireless Bridge Monitoring System Based on MOAP,TP274
  21. Campus network multimedia information system design and development,TP311.52

CLC: > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Information processing (information processing) > Pattern Recognition and devices > Image recognition device
© 2012 www.DissertationTopic.Net  Mobile