Skip to Main Content
This paper presents a new classification method based on a combination of GIS and BP (back propagation) artificial neural network, taking the TM image of the area of Jinzhou city in 2000 as the testing one. BP neural network can be optimized by means of selecting GIS data aided training sample, improving training algorithm, calculating the number of nodes in the hidden layer and so on. Compared with traditional methods, optimized BP neural network is an effective classification method with better self-learning, generalization, adaptive capacity and high classification accuracy.
Geoinformatics, 2010 18th International Conference on
Date of Conference: 18-20 June 2010