Dissertation > Excellent graduate degree dissertation topics show

Content Based Refinement on Video Search Result

Author: ZhangLu
Tutor: ZhouHeQin
School: University of Science and Technology of China
Course: Pattern Recognition and Intelligent Systems
Keywords: Video search results optimization Reordering Content-based video search Particle Swarm Optimization Video Quality Assessment Copy detection As early stop strategy
CLC: TP391.41
Type: Master's thesis
Year: 2010
Downloads: 165
Quote: 1
Read: Download Dissertation

Abstract


With the popularity of digital video equipment, and the development of multimedia technology, the amount of video data on the Internet are growing rapidly. How to effectively store, organize and manage video data on the Internet, has become the main topic of the video field. Major video search engine strategy to provide text-based video search service, such as Google Video Search, Yahoo Video Search, the Bing Video Search and Baidu video search. The basic principles of the algorithm is to get the information of the surrounding text of video on the Internet, then use the text search framework to deal with these information, so as to realize the video searching and sorting. Text-based video search strategies exist, however, the following drawbacks: a large number of the contents of the video data and the surrounding text information inconsistency; video data contained in the complex the amount of visual and auditory information words can not fully express. In order to solve the defects of text-based video search, academics proposed to optimize the search results. The video search results optimized on the basis of the result of the text search, through the excavation of the visual content of the video data contains information on the reordering of the original search results, to thereby obtain a better search results, its fundamental purpose is to improve the search results user experience in quality as well as the search process. Three levels of content-based video search results optimization, the correlation reordering, video quality assessment, and video copy detection. These three levels form a relatively complete framework to optimize video search results: (1) in order to optimize the overall relevance of the original results, reordering Adaptive PSO-based video search and depth analysis of the nature of the problem of reordering. Traditional reordering methods, the paper will be reordering is defined as the process of the evolution of a group, and make full use of the knowledge to guide the direction of the group evolution of each individual to learn to evolution. (2) In order to control the quality of the original search results, this paper based on the content of the video quality assessment and analysis of several key factors that affect the the video overall visual quality. And the traditional signal loss reduction theory video quality assessment system, the assessment system is given in this paper do not need a reference source video. (3) reduce redundant video search results in order to remove the video search results in a large number of duplicate video video copy detection based on the combined features of space-time. This paper analyzes the main types of Internet video copy, and pointed out the need to ensure the efficiency of the detection algorithm to the algorithm is applied in practice. The algorithm uses a combination of rough match and exact match two matching framework early stop strategy, and are joined at each level, and at the same time ensure the accuracy of the detection, greatly improve the detection speed.

Related Dissertations

  1. Research on Feature Extraction and Classification of Tongue Shape and Tooth-Marked Tongue in TCM Tongue Diagnosis,TP391.41
  2. Active Power Filter and Its Application in Distribution Network,TN713.8
  3. Segmentation of cDNA Microarray Image Using Fuzzy C-means Algorithm Optimized by Particle Swarm,TP391.41
  4. Research on Mobile Robot Path Planning and Simulation Realization,TP242
  5. Based on intelligent algorithm research dimensional cutting stock problem,TP301.6
  6. The Research of Explosion Search Algorithm,TP301.6
  7. Research of Variable Neighborhood Search and Application in Combinatorial Optimization,TP301.6
  8. In-furnace Temperature Information Included Combustion Optimization of a Utility Boiler,TK227.1
  9. Research on Multi-time Period Production and Procurement Plan of Supply Chain under Uncertainty,F224
  10. Multi-step-ahead Stock Price Index Forecasting Based on Hybrid Models,F224
  11. Intelligent Support Vector Machine Approach and Its Application on Melt Index Prediction in Propylene Polymerization,TQ325.14
  12. Flexible path based on hybrid particle swarm cross-cell scheduling method,TH165
  13. Genetic hybrid algorithm based on particle swarm Distribution Network Reconfiguration,TM732
  14. Support Vector Machine and Its Application in Burst Prediction Model,TD324
  15. The Research on Intelligent Methods of Power Transformer Fault Diagnosis,TM41
  16. Short-term Power Load Forecasting Based on the Particle Swarm Optimization Algorithm,TM715
  17. Theapplication of Particle Swarm Optimization in Query Optimization,TP391.3
  18. Application of Wavelet Network in Economic Prediction,F201
  19. The hydrometallurgy electrolysis process energy consumption optimization Control Research and application,TF813
  20. Study on Distribution Network Optimization Planning Based on Particle Swarm Optimization Algorithm with Adaptive Velocity,TM715
  21. Improved Particle Swarm Optimization Algorithm under Dynamic Environment,TP18

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