Dissertation > Excellent graduate degree dissertation topics show

Image Segmentation Based on Multi-Agent

Author: FengBo
Tutor: LiuFang
School: Xi'an University of Electronic Science and Technology
Course: Applied Computer Technology
Keywords: Multi-Agent Autonomous behavior Edge amendment operator Image segmentation
CLC: TP391.41
Type: Master's thesis
Year: 2009
Downloads: 43
Quote: 0
Read: Download Dissertation

Abstract


Image segmentation refers to the image is divided into regions with the different characteristics and extract the target of interest in technology and process. It is a key step from the image processing into the image analysis, and the basis of further image understanding. Image segmentation technology is closely related to many other subjects. With the various disciplines of the new theories and methods proposed, many segmentation technology combining the relevant theories, methods and tools appears. Image segmentation based on Multi-Agent is one of them.In the paper that adaptive image segmentation with distributed behavior-based Agents proposed by Jiming Liu, author designs Agents with the autonomous behavior for image segmentation problem. Through the breeding, diffusion of Agents gradually labels homogeneous regions, the eventual realization of image segmentation. It has achieved good results, but there still exist some problems, so this dissertation has made improvements in the following areas:1) For each pixel, its feature is the vector combined gray and gray co-occurrence matrix statistics component. It makes algorithm more applicable to texture image and S AR (Synthetic Aperture Radar) image.2) Initial Agents are classified by its feature similarity. Agents with more similar get together. The algorithm convergence rate is raised than the original initial random distribution.3) The standard of homogeneous region is defined by the adaptive evolution of agents. The algorithm is more applicable, robust and practical than the pre-defined standards in the original paper.4) A competitive operator is proposed. The competitive operator works when two kinds of agents label a pixel at the same time.Experimental results show better performance.There is a problem in image segmentation. That is contradiction that regional consistency and accuracy of the edge of the division. In this dissertation, because of the thinking of agents and the lack of information in spatial domain, an edge detection and amendment method based on transform domain is proposed. The algorithm detects edges using transform domain-based feature firstly, and then applies edge amendment operator. The operator acting on the edge, it takes the characteristics of edge’s shape as a priori knowledge. Pixels near the edge are segmented according to that. The test results show the effectiveness of the method.

Related Dissertations

  1. Research on Image Recognition Algorithm in the Forest Fire Prevention System,TP391.41
  2. Application Research of Digital Image Processing on Container Inspection,TP274.4
  3. Study on the Design and Biomechanical Behaviors of Individualized Artificial Knee Joint,R318.1
  4. Based on statistical methods of magnetic resonance imaging of the human brain image segmentation and three-dimensional analysis of the data,R445.2
  5. Research and Implementation on Service-Oriented Multi-Agent System Cooperation Mechanism,TP393.09
  6. Road extraction algorithm based on region segmentation of remote sensing image,TP751
  7. Tobacco Diseases Auto-Recognition Research Based on Image Processing Technology,S435.72
  8. Research for Infrared Image Target Identification and Tracking Technology,TP391.41
  9. BF-FCM Clustering Algorithm and Its Application in the Image Segmentation,TP391.41
  10. Segmentation of cDNA Microarray Image Using Fuzzy C-means Algorithm Optimized by Particle Swarm,TP391.41
  11. Research on Modeling and Simulation for Credit Risk management and Control System of Bank Based on Complex Multi-Agent Systems,F832.4
  12. Research on Multi-Agent-Based Fire Scene Evacuation Simulation at High-Rised Building,TU972.4
  13. The Study of Human Face Detection Method Based on OpenCV,TP391.41
  14. Computer-assisted hip replacement surgery systems research and development,TP391.41
  15. The Depth Estimation Based on Monocular Night Vision Image,TP391.41
  16. Application of Multi-Agent Methods in Distributed Electric Power Generation Scheduling of Smart Grid,TM76;TM73
  17. Jade Multi-Agent Based Image Retrieval System,TP391.3
  18. Ontology-based Multi-Agent Systems Trading Partner Intelligence research findings related technologies,F713.36
  19. Machining complex manufacturing systems - heat integrated scheduling method,TH186
  20. Based on Active Contour Models Cardiac MRI Segmentation of the left ventricle,TP391.41
  21. Workshop production scheduling based on clustering virtual alliance negotiation mechanism,TP301.6

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