Dissertation > Excellent graduate degree dissertation topics show
Research on Visual Image-based 3D Model Retrieval and Semantic Automatic Annotation Technology
Author: YangShaoBo
Tutor: GengGuoHua
School: Northwestern University
Course: Applied Computer Technology
Keywords: 3D model retrieval Orthogonal projection Perspective projection Semantic annotation
CLC: TP391.41
Type: Master's thesis
Year: 2010
Downloads: 158
Quote: 1
Read: Download Dissertation
Abstract
|
Semantic retrieval is a hot research field of 3D model retrieval , this paper two important aspects of the semantic retrieval feature extraction and semantic annotation technology research , computer vision, image taking into account the three-dimensional model of the most characteristic content features , it is This paper focuses on three-dimensional model of the visual image retrieval and semantic annotation technology - based . Respectively through the orthogonal projection of an image matching computer vision and a perspective projection image to obtain similarity of the three-dimensional model , and the algorithm is applied to the semantic autolabels technology among . The main research work as follows: (1) of the orthogonal projection of the image based on visual model retrieval algorithm . Obtain a three - dimensional model on the basis of the two - dimensional orthogonal projection image , respectively, by comparable projection image contour edge angle and statistical projection image Cylindrical histogram to calculate the image similarity , and thus to obtain a three - dimensional model of similarity , and Both two - dimensional shape feature extraction thinking of expanding to a three-dimensional field . Experimental comparison shows that such algorithm has better retrieval efficiency , and for 3D model retrieval request translation, rotation and scaling invariance has better robustness . (2) based on 6 - viewpoint and a 14 - point of view perspective projection image model retrieval algorithm . Similarity to calculate the projection image using the contour edge angle based on the projection image , and a method based on the projection image Cylindrical experimental comparison shows that the viewpoint would increase the retrieval accuracy of the model under specific projection matching algorithm . (3) proposed an improved semantic tagging algorithm automatically . Combined with the proposed computing model retrieval algorithm based on orthogonal projection image edge angle and Jiang Conrath semantic vocabulary similarity algorithm , the model automatically marked . The algorithm effectively improve model Find and semantic vocabulary matching accuracy based on the improvement of the effect of the automatic annotation . (4) system . Design and development of a three-dimensional model of browsing, editing and retrieval retrieval prototype system , the system validates this article involves the retrieval algorithm , has a simple , interactive , scalability . This article National 863 High-Tech Research and Development Program (2008AA012301) - three-dimensional model of intelligent processing and retrieval platform support .
|
Related Dissertations
- The Research on Electrode 3D Model Classification and Retrieval Based on SVM and Shape Features,TP391.41
- Research and Implementation on Sns-based Fine-grained Annotation of Educational Video,TP391.6
- Image semantic annotation of blocks - a global feature extraction method,TP391.41
- Weak sparse underdetermined blind signal separation technology research,TN911.7
- Human motion sequence data semantic analysis method,TP391.1
- Research on the Semantic Annotation of Domain with WordNet,TP391.1
- The Research of Data Extraction and Semantic Annotation in Deep Web,TP393.09
- Camera-based network monitoring system,TP277
- 3D Mesh Segmentation Based on the Strength of Boundary,TP391.41
- Research on Constrained Matrix Equations and Their Preconditioning Techniques of Iterative Methods,O241.6
- Application of Model Segmentation in Content-based 3d Model Retrieval,TP391.41
- Dynamic real-time detection field optical CT reconstruction algorithm design and implementation,TP391.41
- 3D model retrieval algorithm based on local shape distribution,TP391.41
- Researches on 3d Model Retrieval Based on XML Technology and Semantic Matrix,TP391.41
- A Note on Nonlinear Kalman Filter,O211.64
- Exponential Attractor for Coupled Ginzburg-Landau Equations,O175.29
- Research and Application of Intelligent Search Engine Based on Ontology,TP391.3
- Design and Implementation of Semantic Annotation System for 3D Models,TP391.41
- Research on Statistics and Evaluation Methods of Railway Emergency Management,U298
- Research on 3D Model Acquisition and Auto-tagging Algorithm,TP391.41
- The Study of Semantic Annotation Technique of Insects Image,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
|