An Improved Heuristic Optimization Algorithm for Feature Learning Based on Morphological Filtering and its Application

IEEE Access

卷:6

Shengjun Chen¹  Chengming Qi²  Qun Wang³  Lishuan Hu³ 

[1] Business School, University of International Business and Economics, Beijing, China;

[2] College of Urban Rail Transit and Logistics, Beijing Union University, Beijing, China;

[3] School of Information Engineering, China University of Geosciences, Beijing, China;

关键词: Binary particle swarm optimization;    feature learning;    hyperspectral image classification;    mathematical morphology;   

DOI  :  10.1109/ACCESS.2018.2827403

来源: DOAJ

【 摘 要 】

Hyperspectral remote sensing sensors can provide plenty of valuable information with hundreds of spectral bands at each pixel. Feature selection and spectral-spatial information play an important role in the field of hyperspectral image (HSI) classification. In this paper, a novel two-stage spectral-spatial HSI classification method is proposed. In first stage, the standard particle swarm optimization (PSO) is adopted to optimize the parameters, and a novel binary PSO with mutation mechanism is used for feature selection simultaneously. Then, the support vector machine classifier is performed. In second stage, in order to reduce salt and pepper phenomenon, mathematical morphology post-processing is used to further refine the obtained results of the above stage. Experiments are conducted on two real hyperspectral data sets. The evaluation results show that the proposed approach achieves better accuracy than several state-of-the-art methods.

【 授权许可】

Unknown