Data processing: database and file management or data structures – Database design – Data structure types
Reexamination Certificate
1999-01-26
2002-09-03
Vu, Kim (Department: 2172)
Data processing: database and file management or data structures
Database design
Data structure types
C707S793000
Reexamination Certificate
active
06446060
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates generally to improved information retrieval systems for images and other nonstructure data, and more particularly to converting a query with fuzzy specifications of one or more objects and spatial or temporal relationships between different objects into a set of subgoals, organizing the entire search space into blocks blocks according to the set of subgoals; and processing the query blocks in the order of likelihood of satisfying, the final constraints without possible false dismissal.
BACKGROUND OF THE INVENTION
It is becoming increasingly important for multimedia databases to provide capabilities for content-based retrieval of composite objects. Composite objects consist of several simple objects which have feature, spatial, temporal, semantic attributes, and spatial and temporal relationships between them. Recent methods for retrieving images and videos by content from large archives utilize feature descriptions and feature comparison metrics in order to index the visual information. Examples of such content-based retrieval systems include the IBM Query by Image Content (QBIC) system detailed in “Query by image and video content: The {IQBI} system.” by M. Flickner, et al in “IEEE Computer”, 28(9):23-32. (September 1995); the Virage visual information retrieval system detailed in “Virage image search engine: an open framework for image management,” by J. Bach, et al in “Symposium on Electronic Imaging: Science and Technology—Storage & Retrieval for Image and Video Databases {IV}”, volume 2670, pages 76-87. (1996), the MIT Photobook, detailed in “Tools for content-based manipulation of image databases,” by A. Pentland, et al in “Proceedings of the SPIE Storage and Retrieval Image and Video Databases II”, (February 1994); the Alexandria project at UCSB detailed in “Texture features for browsing and retrieval of image data,” by B. S. Manjunath and W. Y. Ma, “IEEE Trans. Pattern Analysis Machine Intell. Special Issue on Digital Libraries”, 8 (1996) and in “Dimensionality reduction using multidimensional scaling for image search,” by M. Beatty and B. S. Manjunath, published in the “Proc. IEEE International Conference on Image Processing” (October 1997); and the IBM/NASA Satellite Image Retrieval System detailed “Progressive content-based retrieval from distributed image/video databases,” by in C.-S. Li, V. Castelli, and L. Bergman in the “Proceedings of the International Symposium of Circuit and System”, IEEE (1997).
It is becoming increasingly important for multimedia databases to provide capabilities for content-based retrieval of composite objects. Composite objects consist of several simple objects which have features, spatial, temporal and semantic attributes, and spatial and temporal relationships between them. The need for compound search queries frequently arises in various scientific and engineering applications, for processing such as the following:
Retrieve those Synthetic Aperture Radar (SAR) Satellite images and identify those regions in the images with texture type (e.g., ice) similar to the search target,
Retrieve those one-meter resolution satellite images and identify those regions in the images with spectral features (e.g., crop) similar to the search target,
Retrieve those LANDSAT Thematic Mapper (TM) satellite images and identify those regions in the images with a combination of spectral and texture features (e.g., indicative of terrain type) similar to the search target.
These scenarios frequently arise in the following applications:
Environmental epidemiology: wherein there is a need to retrieve locations of houses which are vulnerable to epidemic diseases such as Hantavirus and Denge fever based on a combination of environmental factors (e.g. isolated houses that are near bushes or wetlands), and weather patterns (e.g. a wet summer followed by a dry summer);
Precision farming: for (1) retrieving locations of cauliflower crop developments that are exposed to clubroot, which is a soil-borne disease that infects cauliflower crop, where cauliflower and clubroot are recognized spectral signature, and exposure results from their spatial and temporal proximity; and (2) retrieving those fields which have abnormal irrigation, (3) Retrieve those regions which have higher than normal soil temperature;
Precision forestry: for (1) calculating areas of forests that have been damaged by hurricane, forest fire, or storms, and (2) estimating the amount of the yield of a particular forest;
Petroleum exploration: where there is a need to retrieve those regions which exemplify specific characteristics in the collection of seismic data, core images, and other sensory data;
Insurance: for (1) retrieving those regions which may require immediate attention due to natural disasters such as earthquake, forest fire, hurricane, and tornadoes, and (2) retrieving those regions have higher than normal claim rate (or amount) that are correlated to the geography—close to coastal regions, close to mountains, in high crime rate regions, etc.;
Medical image diagnosis: for retrieval of all MRI images of brains that have tumors located within the hypothalamus, which tumors are characterized by shape and texture, and the hypothalamus is characterized by shape and spatial location within the brain;
Real estate marketing: to retrieve all houses that are near a lake (color and texture), have a wooded yard (texture) and are within 100 miles of skiing (mountains are also given by texture); and
Interior design: for retrieval of all images of patterned carpets which consist of a specific spatial arrangement of color and texture primitives.
Until recently, content-based query and spatial query paradigms have been largely distinct. On one hand, there has been extensive investigation of using logical representations to facilitate efficient processing of spatial and temporal queries of symbolic images (see: “Design and Evaluation of Algorithms for Image Retrieval by Spatial Similarity”, by V. N. Gudivada and V. V. Raghavan, in the “ACM Trans. on Information Systems”, Vol. 13, No. 2 (April 1995); “Iconic Indexing by {2-D} Strings”, by S.-K. Chang and Q. Y. Shi and C. Y. Yan, “IEEE Trans. on Pattern Recognition and Machine Intelligence”, Vol. 9, No. 3, pp. 413-428 (May 1987)), and videos (“OVID: Design and Implementation on a Video-Object Database System”, by E. Oomoto and K. Tanaka, “Trans. on Knowledge and Data Engineering”, vol. 5 (August 1993)). The various logical representations such as 2D-strings, the &THgr;-R representation, and the spatial orientation graph (SOG) allow indexing and retrieval based upon spatial and temporal relationships.
On the other hand, content-based image retrieval systems, such as Virage, Photobook and QBIC, allow querying based upon image features. Examples of the visual features supported by these systems include color, texture, shape, edge, and so forth.
The integration of content-based and spatial image query methods is only recently begun to be investigated (see the above-referenced publications of Li et al, Smith and Chang, and, “Retrieval by content in symbolic-image databases”, by A. Soffer and H. Samet, in the “Symposium on Electronic Imaging: Science and Technology—Storage & Retrieval for Image and Video Databases IV″, 2670, pp144-155 (1996)”; “Content-Based Indexing of Spatial Objects in Digital Libraries”, by S.-F Chang et al; S.-S. Chen, in the “Journal of Visual Communication and Image Representation”, 1, pp.16-27 (March 1996); E, “Similarity Searching in Large Image Databases”, by. G. M. Petrakis and C. Faloutsos, “Department of Computer Science, University of Maryland”, 3388, (1995).) The SaFe spatial and feature image query engine computes the spatial and feature queries by decomposing the composite query into parallel, content-based region queries. After the joining the results, the spatial relationships are evaluated only for the surviving images using query-time 2D-string projection and comparison. SaFe is also being extended to spatial and temporal querying of video (s
Bergman Lawerence D.
Castelli Vittorio
Li Chung-Sheng
Smith John R.
Dougher Anne Vachon
International Business Machines - Corporation
Jordan Kevin M.
Nguyen Tam V
Vu Kim
LandOfFree
System and method for sequential processing for... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with System and method for sequential processing for..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and System and method for sequential processing for... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2859537