3D Content Retrieval

3D Content Retrieval system is a computer system for browsing, searching and retrieving three dimensional digital contents (eg: Computer-aided design, molecular biology, and cultural heritage 3D scenes, etc.) from a large database of digital images. The most original way of doing 3D content retrieval uses methods to add text to 3D content files. Because of the inefficiency of manually annotating 3D files, researchers have investigated ways to automate the annotation process and provide a unified standard to create text descriptions for 3D content. Moreover, the increase in 3D information. Thus, shape matching methods for 3D content retrieval have become popular. Shape matching is based on techniques that compare and contrast similarities between 3D models.

3D retrieval methods

High level description (eg: skeleton) and then find matching results

This method describes 3D models by using a skeleton. The skeleton encodes the geometric and topological information in the form of a skeletal graph and uses graph matching techniques to match the skeletons and compares them. [1] However, this method requires a 2-manifold input model, and it is very sensitive to noise and details. Many of the existing 3D models are created for visualization purposes, while missing the standard input for the skeleton method. The skeleton 3D retrieval method needs more time and effort before it can be used widely.

Compute a feature vector based on statistics

Skeleton modeling, which requires a high quality standard for the input source. Shape histograms , feature vectors consisting of global geo-metic properties such as circularity and eccentricity, and feature vectors using the method of decomposition of spherical functions. [2]

2D projection method

Some approaches use 2D projections of a 3D model, but they have similar 2D projections in many directions. Prototypical Views [3] and Light field Description [4] are good examples of 2D projection methods.

3D Engineering Search System

In Purdue University, researchers led by Professor Karthik Ramani at the Research and Education Center for Information created a 3D search engine called the 3D Engineering Search System (3DESS). It is designed to find computer-generated engineering parts.

The mechanism behind this search engine is That It starts from an algorithm qui can transform query drawing to voxels , Then extracts the significant MOST Information shape from the voxels by using Reviews another algorithm called Expired thinning, and formulates a skeleton of the object’s outlines and topology . After that, 3DESS will develop a skeletal graph to render the skeleton, using three common topological constructs: loops, edges, and nodes. The processed common constructs can be represented in the database. [5]

According to the lead professor, 3DESS can also describe objects using feature vectors, such as volume, surface area, etc. The system processes queries by their data vectors or skeletal graphs with data stored in the database. When the system retrieves models in response to the query, users can pick whichever object.

Challenges

Challenges associated with 3D shape-based similarity queries

With the skeleton 3D modeling retrieval method, figuring out an efficient way to 3D 3D shape indexing is very challenging because 3D shape indexing has very strict criteria. The computational model, which is easy to index, invariant under similarity transformations, insensitive to noise and small features, robust to arbitrary topological degeneracies, and discrimination of shape differences.

3D search and recovery with multimodal support challenges

In the search for a simple search engine, a multimodal retrieval system, which can take various types of input sources and provide robust query results, is necessary. So far, only a few approaches have been proposed. In Funkhouser et al. (2003), [6] the proposed “Princeton 3D search engine” supports 2D sketches, 3D sketches, 3D models and text as queries. In Chen et al. (2003), [7] he has designed a 3D retrieval system that incorporates 2D sketches and retrieves for 3D objects. Recently, Ansary et al. (2007) [8] proposed a 3D retrieval framework using 2D photographic images, sketches, and 3D models.

References

  1. Jump up^ Sundar, H., Silver, D., Gagvani, N., Dickinson, S., Skeleton based shape matching and retrieval, In: Proc. SMI, Seoul, Korea (2003)
  2. Jump up^ Min, P., Kazhdan, M., Funkhouser, T., A comparison of text and shape matching for retrieval of Online 3D models. Research And Advanced Technology For Digital Libraries, 2004, Vol.3232, pp.209-220
  3. Jump up^ Cyr, CM, Kimia, BB, 3D object recognition using shape similarity-based appearance graph, In: Proc. ICCV, IEEE (2001)
  4. Jump up^ Chen, DY, Ouhyoung, Mr. Tian, XP, Shen, YT, Ouhyoung, M., visual similarity is based 3D model retrieval, In: Proc. Eurographics, Granada, Spain (2003)
  5. Jump up^ Ortiz, S., 3D searching starts to take shape, Computer, 2004, Vol.37 (8), pp.24-26
  6. Jump up^ Funkhouser, T., Min, P., Kazhdan, M., Chen, J., Halderman, A., Dobkin, D., & Jacobs, D. (2003). A search engine for 3D models. ACM Transactions on Graphics, 22 (1), 83-105
  7. Jump up^ Chen, DY, Ouhyoung, Mr. Tian, XP, Shen, YT, Ouhyoung, M., visual similarity is based 3D model retrieval, In: Proc. Eurographics, Granada, Spain (2003)
  8. Jump up^ Filali Ansary, T., Daoudi, M., & Vandeborre, J.-P. (2007). A 3D search engine Bayesian using adaptive views clustering. IEEE Transactions on Multimedia, 9 (1), 78-88.

Leave a Reply

Your email address will not be published. Required fields are marked *

Copyright computerforum.eu 2018
Shale theme by Siteturner