Statistics – Applications
Scientific paper
Sep 2004
adsabs.harvard.edu/cgi-bin/nph-data_query?bibcode=2004esasp.553e..21d&link_type=abstract
Proceedings of ESA-EUSC 2004 - Theory and Applications of Knowledge-Driven Image Information Mining with Focus on Earth Observa
Statistics
Applications
Image Information Mining, Cbir, System Evaluation, Hmi
Scientific paper
In this article, we present tools for the evaluation of a knowledge-drivencontent-based image information mining system. In order to provide users fast access to the content of large remote sensing image archives, the system is composed of two main modules. The first includes computationally intensive algorithms for off-line data ingestion in the database, feature extraction and indexing. The second module consists of a graphical human-machine interface that manages the interactive learning and image information mining functions. According to the system architecture, the implemented evaluation tools determine the objective technical quality of the system and include subjective human factors, too. Since the query performance of the mining system mainly depends on the data sets stored in the archive, we first analyze the complexityof image data. Based on the stochastic nature of user-defined semantic cover-type labels, the system retrieves the most relevant images using probabilistic measurements. We evaluate the man-machine communication dialogue and system operation in order to determine the quality of semantic labels. Finally, we verify the man-machine interfaceby using measurements like time for loading the learning applet, time for computing the probabilistic search results and time for label training.
Daschiel H.
Datcu Mihai
No associations
LandOfFree
Evaluation Tools for Image Information Mining System does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.
If you have personal experience with Evaluation Tools for Image Information Mining System, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Evaluation Tools for Image Information Mining System will most certainly appreciate the feedback.
Profile ID: LFWR-SCP-O-1064418