Neurofuzzy Analysis of Cloud Motion from Time Series of Satellite Imagery s.n.
Full text not available from this repository.
This report outlines the first year s research into the application of neural networks and fuzzy logic to the extraction of wind motion vectors from time series of satellite imagery. To improve on current schemes for determining wind information from cloud motion, three new components of satellite image sequences have been used: context, cloud 'object' motion analysis and longer time series. A fuzzy system has been used to identify the motion context within an image sequence of 12 hours of satellite data. Techniques for the automatic extraction of cloud objects and regions have been looked at, and an array clustering algorithm defined on interest operators in an image sequence of 4 hours of data has been developed. A new neurofuzzy approach to cloud region identification and motion analysis is being developed for application to trial images supplied with operationally generated wind data, and is outlined in this report. A timetable for the remainder of the studentship is given.
||Address: Department of Electronics and Computer Science
||Electronics & Computer Science
||04 May 1999
||18 Apr 2017 00:24
|Further Information:||Google Scholar|
Actions (login required)