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
- Operational Risk of Commercial Banks in China,F832.2
- Automatic Image Annotation and Fast Similarity Search in Image Retrieval,TP391.3
- Research on Web Image Retrieval Based on the Fusion of Textual and Visual Information,TP391.3
- Visual Language Analysis: From Low Level Feature Representation to Semantic Distance Learning,TP391.41
- Research on Real-World Automatic Image Annotations,TP391.41
- Research on Perception Oriented Image Retrieval and Automatic Image Annotation,TP391.3
- Internet-based image annotation dataset Research,TP391.41
- Internet environment large image content analysis , retrieval and automatic labeling of,TP391.41
- Research on Key Techniques of Automatic Image Annotation,TP391.41
- Research on Image Retrieval with Machine Learning Techniques,TP391.41
- Research on Image Annotation and Image Annotation Refinement,TP391.41
- Some Results of Modeling and Learning Genetic Regulatory Networks,Q78
- Research of Large-Scale Web Image Annotation and Interpretation,TP391.41
- Research on Key Techniques of User Oriented Web Image Retrieval,TP391.41
- Image Mining in Image Retrieval,TP391.41
- Selective Oxidation of Coals in Aqueous NaOCl Solution and Fine Separation of the Products,TQ533
- Research on Automatic Annotation in Image Retrieval,TP391.41
- Research on Technologies of Image Semantic Annotation and Retrieval,TP391.41
- An Ontology-Based Image Retrieval System,TP391.3
- 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
|