Fayyad, U. and Smyth, P. Image Database Exploration: Progress and Challenges. In Proceedings of 1993.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....be fairly unstable with accuracy varying from 30 to 95 . About 5 x l0 s objects were classified. Obtained reso lution was one magnitude better than in the previous astronomical studies and it was possible to classify objects with images too faint to be classified by astronomers. Magellan Study [3] analyzed about 30,000 high resolution radar images of the surface of Venus. The goal was to identify volcanos, a task which would take about 10 man years if performed manually. The system is composed of three basic components: data focusing, feature extraction, and classification learning. Like ....

....plate classification. Dealing with Uncertainty. In general, it is difficult for experts to provide classifications with 100 certainty and false classifications can produce large errors during classification because they are treated as negative examples. Smyth et al. (in [4] used Magellan Study [3] to analyze such issues. The work is on the modeling and treatment of subjective label information given by the experts using probabilistic models. It shows that it is possible for the knowledge discovery methods to be modified to handle the lack of absolute ground truths. Scalable Mining of ....

Fayyad, U. and Smyth, P. Image Database Exploration: Progress and Challenges. In Proceedings of 1993.

....parks . 2.3 Other Relevant Studies Also knowledge mining in image databases, which can be treated as a special type of spatial databases, has been studied recently. Method for the classification of sky objects and another method for recogntion of volcanos on the surface of Venus are described in [8], where classification trees were used to make final decisions. Sky objects were classified as stars or galaxies. In the first step of the algorithm, basic attributes describing each object were extracted. Attributes like area, sky brightness, positions of peak brightness, and intensity image ....

U. Fayyad, and P. Smyth. Image Database Exploration: Progress and Challenges. In Proc. 1993 Knowledge Discovery in Databases Workshop, pp. 14-27, Washington, D.C..

....[13] aimed at the retrieval of raster data, while the Sloan Digital Sky Survey [14] poses the need for the creation of multiterabyte astronomy archive. The large scale systems for the analysis of remotely sensed images were specialized toward the detection of particular features like volcanoes [2], or proposed distributed and parallel data storage and query processing systems for handling of geo scientific data retrieval queries [11] The GeoBrowse project aims to provide infrastructure that would enable the analysis of large databases containing satellite images. Our work addresses two ....

Fayyad, U. M., and P. Smyth. Image Database Exploration: Progress and Challenges. In Proc. 1993 Knowledge Discovery in Databases Workshop, Washington, DC, p. 14 -- 27, 1993.

....this, as can be expected, is that there are a number of issues coming forth from these fields of studies that relate to the progress and development of spatial data mining. There has been a considerable number of studies being done in machine learning (for e.g. 38, 39, 16] databases (for e.g. [26, 1, 14]) and statistics (for e.g. 18, 32, 10] that somewhat lay the foundation for knowledge discovery in spatial databases. Development of data mining is challenging to every field discussed above, escpecially due to the fact that the data considered in this case is very large, upto several ....

....idea to do so because it will make it clear whether the algorithm is efficient and scalable as it is claimed to be. 10 2.5 Pattern Recognition Method Knowledge mining from Image databases can be said to be a special case of spatial data mining. There has been some studies, led by Fayyad et al. [14, 45], on the automatic recognition and catagorization of sky objects, like volcanoes, in Venus by processing the pictures taken by Magellan spacecraft. Their method is based upon pattern recognition and machine learning techniques. Magellan transmitted more than 30,000 high resolution images of the ....

[Article contains additional citation context not shown here]

U. M. Fayyad and P. Smyth. Image database exploration: Progress and challenges. Proceedings of AAAI-93 workshop on KDD, 1993.

....unstable. Additional drawback of neural networks was the requirement to specify internal parameter such as the number of hidden layers or size. For future investigation, testing of unsupervised clustering techniques is planned. 2. 7 Other Methods The Decision Tree problem raised by Fayyad et al. [9, 10] was followed up by Bell et al. [13] who proposed a method for knowledge discovery in spatial databases based upon evidence theory [14] Evidential reasoning [14] is a generalization of conventional probability in the sense that it does not make any assumptions about the independence of data being ....

U. M. Fayyad and P. Smyth. Image Database Exploration: Progress and Challenges. In Proc. 1993 Knowledge Discovery in Databases Workshop, pp. 14-27, Washington, D. C., July 1993.

.... A three dimensional iconic environment for image database querying , IEEE T SE, 19, 10, Oct. 1993, 997 1011. 17] C. Faloutsos, M. Flickner, W. Niblack, D. Petkovic, W. Equitz, R. Barber, Efficient and effective querying by image content , IBM Research Report, RJ 9453 (83074) August 3, 1993. [18] U.M. Fayyad and P. Smyth, Image database exploration: progress and challenges , 1993 AAAI Workshop on Knowledge discovery in databases, July 11 12, 1993, Washington DC, AAAI Press, Menlo Park, CA, Tech. Rep. WS 93 02. 19] O. Faugeras, Three dimensional Computer Vision: A Geometric Viewpoint, ....

....appearing, commercially or within the research community. Regarding static images, applications can be found in the following domains: medical (Picture Archiving and Communication Systems PACS, e.g. 46] geographical (Geographical Information Systems GIS [34] 49] earth space observation [18]) museums and libraries (e.g. 5] 12] 35] 43] 47] security (fingerprints, faces [3] 42] artifact catalogs, home computing (handling of photographs) news agencies, etc. Regarding video sequences, applications now revolve around automated channel selection, surveillance, or movies on ....

U.M. Fayyad and P. Smyth, "Image database exploration: progress and challenges", 1993 AAAI Workshop on Knowledge discovery in databases, July 11-12, 1993, Washington DC, AAAI Press, Menlo Park, CA, Tech. Rep. WS 93-02.

....unstable with accuracy varying from 30 to 95 . About 5 Theta 10 8 objects were classified. Obtained resolution was one magnitude better than in the previous astronomical studies and it was possible to classify objects with images too faint to be classified by astronomers. Magellan Study [3] analyzed about 30,000 high resolution radar images of the surface of Venus. The goal was to identify volcanos, a task which would take about 10 man years if performed manually. The system is composed of three basic components: data focusing, feature extraction, and classification learning. Like ....

....plate classification. Dealing with Uncertainty. In general, it is difficult for experts to provide classifications with 100 certainty and false classifications can produce large errors during classification because they are treated as negative examples. Smyth et al. in [4] used Magellan Study [3] to analyze such issues. The work is on the modeling and treatment of subjective label information given by the experts using probabilistic models. It shows that it is possible for the knowledge discovery methods to be modified to handle the lack of absolute ground truths. Scalable Mining of ....

Fayyad, U. and Smyth, P. Image Database Exploration: Progress and Challenges. In Proceedings of 1993 Knowledge Discovery in Databases Workshop, (July 11-12, Washington, DC). AAAI Press, Menlo Park, CA, 1993, pp. 14-27.

....parks . 2.3 Other Relevant Studies Also knowledge mining in image databases, which can be treated as a special type of spatial databases, has been studied recently. Method for the classification of sky objects and another method for recogntion of volcanos on the surface of Venus are described in [8], where classification trees were used to make final decisions. Sky objects were classified as stars or galaxies. In the first step of the algorithm, basic attributes describing each object were extracted. Attributes like area, sky brightness, positions of peak brightness, and intensity image ....

U. Fayyad, and P. Smyth. Image Database Exploration: Progress and Challenges. In Proc. 1993 Knowledge Discovery in Databases Workshop, pp. 14-27, Washington, D.C..

.... A three dimensional iconic environment for image database querying , IEEE T SE, 19, 10, Oct. 1993, 997 1011. 17] C. Faloutsos, M. Flickner, W. Niblack, D. Petkovic, W. Equitz, R. Barber, Efficient and effective querying by image content , IBM Research Report, RJ 9453 (83074) August 3, 1993. [18] U.M. Fayyad and P. Smyth, Image database exploration: progress and challenges , 1993 AAAI Workshop on Knowledge discovery in databases, July 11 12, 1993, Washington DC, AAAI Press, Menlo Park, CA, Tech. Rep. WS 93 02. 19] O. Faugeras, Three dimensional Computer Vision: A Geometric Viewpoint, ....

....appearing, commercially or within the research community. Regarding static images, applications can be found in the following domains: medical (Picture Archiving and Communication Systems PACS, e.g. 46] geographical (Geographical Information Systems GIS [34] 49] earth space observation [18]) museums and libraries (e.g. 5] 12] 35] 43] 47] security (fingerprints, faces [3] 42] artifact catalogs, home computing (handling of photographs) news agencies, etc. Regarding video sequences, applications now revolve around automated channel selection, surveillance, or movies on ....

U.M. Fayyad and P. Smyth, "Image database exploration: progress and challenges", 1993 AAAI Workshop on Knowledge discovery in databases, July 11-12, 1993, Washington DC, AAAI Press, Menlo Park, CA, Tech. Rep. WS 93-02.

....to be scalable and the memoization policy is found to be the most consistent and efficient of all the shape enlargement policies. 2.5 Mining in Image Databases Knowledge mining from Image Databases can be viewed as a case of spatial data mining. There have been studies, led by Fayyad et al. [14, 15, 48], on the automatic recognition and categorization of astronomical objects. The authors presented a system [15] for identifying volcanos on the surface of Venus from images transmitted by the Magellan spacecraft. The Magellan transmitted more than 30,000 high resolution synthetic aperture radar ....

....policies. 2.5 Mining in Image Databases Knowledge mining from Image Databases can be viewed as a case of spatial data mining. There have been studies, led by Fayyad et al. 14, 15, 48] on the automatic recognition and categorization of astronomical objects. The authors presented a system [15] for identifying volcanos on the surface of Venus from images transmitted by the Magellan spacecraft. The Magellan transmitted more than 30,000 high resolution synthetic aperture radar images of the surface of Venus from different angles. The system is composed of three basic compo11 spatial data ....

[Article contains additional citation context not shown here]

U. M. Fayyad and P. Smyth. Image Database Exploration: Progress and Challenges. In Proc. 1993 Knowledge Discovery in Databases Workshop, pp. 1427, Washington, D.C., July 1993.

....volcano classifiers and for comparing the performance of various algorithms and humans. Because there is a wide variation in the labelings created by individual experts, the ground truth used for this study was the consensus of two planetary geologists who discussed and labeled the images together (Fayyad Smyth 1993). This is the ground truth labeling used for most of the JARtool work at JPL. It consists of 163 volcanos in four sample Magellan SAR images. Smyth et al. 1995) describe an alternative, algorithmic approach to estimating ground truth from the individual labelings of many experts. The Current ....

Fayyad, U., and Smyth, P. 1993. Image database exploration: Progress and challenges. In Proc Knowledge Discovery in Databases Wkshp, 11th Nat Conf on Artificial Intelligence. Washington, DC: AAAI Press.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC