Remote-Sensing Image Classification Based on an Improved Probabilistic Neural Network 英文参考文献.docVIP

Sensors 2009, 9, 7516-7539; doi:10.3390/OPEN ACCESS sensors ISSN 1424-8220 /journal/sensors Article Remote-Sensing Image Classification Based on an Improved Probabilistic Neural Network Yudong Zhang 1,*, Lenan Wu 1, Nabil Neggaz 2, Shuihua Wang 1 and Geng Wei 1 1 School of Information Science and Engineering, Southeast University, Nanjing 210009, China; E-Mails: shuihuaw2007@ (S.W.); wei_geng@163.com (G.W.) 2 Signal-Image-Parole Laboratory, Department of Computer Science, University of Science and Technology – Oran, Oran, Algeria; E-Mail: neggaz_nabil@yahoo.fr * Author to whom correspondence should be addressed; E-Mail: zhangyudongnuaa@. Received: 12 June 2009; in revised form: 2 September 2009 / Accepted: 16 September 2009 / Published: 23 September 2009 Abstract: This paper proposes a hybrid classifier for polarimetric SAR images. The feature sets consist of span image, the H/A/α decomposition, and the GLCM-based texture features. Then, a probabilistic neural network (PNN) was adopted for classification, and a novel algorithm proposed to enhance its performance. Principle component analysis (PCA) was chosen to reduce feature dimensions, random division to reduce the number of neurons, and Brent’s search (BS) to find the optimal bias values. The results on San Francisco and Flevoland sites are compared to that using a 3-layer BPNN to demonstrate the validity of our algorithm in terms of confusion matrix and overall accuracy. In addition, the importance of each improvement of the algorithm was proven. Keywords: polarimetric SAR; Probabilistic neural network; gray-level co-occurrence matrix; principle component analysis; Brent’s Search 1. Introduction The classification of different objects, as well as different terrain characteristics, with single channel monopolarisation SAR images can carry a significant amount of error, even when operating after multilooking [1]. One of the m