Dissertation > Excellent graduate degree dissertation topics show

Research and Implementation on Image Annotation Using Probability Modeling

Author: DingLei
Tutor: XuDe
School: Beijing Jiaotong University
Course: Computer Science and Technology
Keywords: Image annotation Probabilistic modeling Related Model
CLC: TP391.41
Type: Master's thesis
Year: 2010
Downloads: 94
Quote: 1
Read: Download Dissertation

Abstract


Automatic image annotation is a solve the artificial Dimensioning challenging work , it tries to build a bridge between high-level semantic features and low - level visual features . With the continuous development of machine learning theory in particular , many scholars have devised different learning models , can be broadly divided into two categories , based on probabilistic modeling image annotation and classifier - based image annotation . The paper studies two representative labeling algorithm based on probabilistic modeling , are the co-occurrence and translation models . Co-occurrence model image is divided into the regular region , to label the image according to the image area and keywords co-occurrence probability , that is, to observe the joint probability of occurrence of keywords and image area . Translation model to improve the co-occurrence model , a new concept to provide a description of the image - visual word element . Visual word element image features clustering , each image contains a visual word set of image annotation can be seen as a process of visual word \The idea of co-occurrence and translation models , we design an improved model . Assume that a labeled training set of images obtained by the image into clustering visual word set , then each picture can be in combination with two sets of visual word element and keywords . Set to give a test image using the language generated model of the methods assume that there is a potential for the probability distribution , i.e. the model , which contains all of the keywords that might appear in the image and visual term , then the annotation process is the probability distribution of a random sampling . Can be approximated by the training set is estimated that the joint distribution , then extracted by the size of the sampling probability value most representative keywords as image annotation results . The improved modeling techniques can effectively utilize large-scale annotated set of training images to achieve better marked effect . Finally, on the Corel dataset experiments confirmed the validity of the model .

Related Dissertations

  1. Operational Risk of Commercial Banks in China,F832.2
  2. Automatic Image Annotation and Fast Similarity Search in Image Retrieval,TP391.3
  3. Research on Web Image Retrieval Based on the Fusion of Textual and Visual Information,TP391.3
  4. Visual Language Analysis: From Low Level Feature Representation to Semantic Distance Learning,TP391.41
  5. Research on Real-World Automatic Image Annotations,TP391.41
  6. Research on Perception Oriented Image Retrieval and Automatic Image Annotation,TP391.3
  7. Internet-based image annotation dataset Research,TP391.41
  8. Internet environment large image content analysis , retrieval and automatic labeling of,TP391.41
  9. Research on Key Techniques of Automatic Image Annotation,TP391.41
  10. Research on Image Retrieval with Machine Learning Techniques,TP391.41
  11. Research on Image Annotation and Image Annotation Refinement,TP391.41
  12. Some Results of Modeling and Learning Genetic Regulatory Networks,Q78
  13. Research of Large-Scale Web Image Annotation and Interpretation,TP391.41
  14. Research on Key Techniques of User Oriented Web Image Retrieval,TP391.41
  15. Image Mining in Image Retrieval,TP391.41
  16. Selective Oxidation of Coals in Aqueous NaOCl Solution and Fine Separation of the Products,TQ533
  17. Research on Automatic Annotation in Image Retrieval,TP391.41
  18. Research on Technologies of Image Semantic Annotation and Retrieval,TP391.41
  19. An Ontology-Based Image Retrieval System,TP391.3
  20. Cross-Media Relevance Models Based Image Annotation and Its Application,TP391.41

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