1. Research Center of Geography and Ecological Environment, Fuzhou University, Fuzhou 350108, China
2. College of Environmental and Safety Engineering, Fuzhou University, Fuzhou 350108, China
3. The Academy of Digital China, Fuzhou University, Fuzhou 350108, China
4. Fujian Provincial Key Laboratory of Resources and Environment Monitoring & Sustainable Management and Utilization, Sanming University, Sanming 365004, China
5. Key Laboratory of Spatial Data Mining & Information Sharing, Ministry of Education, Fuzhou 350108, China
6. Postdoctoral Research Station of Information and Communication Engineering, Fuzhou University, Fuzhou 350108, China
Abstract:Hyperspectral images are characterized by continuous bands, high dimensionality, large data volume and strong correlation between adjacent bands, which can provide richer detailed information for feature classification. However, there is a lot of redundant information and noise in data, and the direct use of all band features without effective analysis and selection in image classification will lead to low computational efficiency and high computational complexity, and the “Hughes phenomenon” that the classification accuracy may increase and then decrease with the increase of band dimension. In order to quickly extract a subset of features with good recognition ability from hyperspectral images with tens or even hundreds of bands to avoid the “dimensional disaster”. This paper combines the filtered ReliefF algorithm and the wrapped recursive feature elimination algorithm (Recursive feature elimination, RFE) to build the ReliefF-RFE feature selection algorithm, which can be used for feature selection in hyperspectral image classification. The algorithm uses the ReliefF algorithm to quickly eliminate many irrelevant features based on weight thresholds to narrow and optimize the range of feature subsets. The RFE algorithm is used to further search for the optimal feature subsets, and the recursive elimination of the less relevant features and redundant to the classifier in the narrowed feature subsets is performed to obtain the feature subsets with the best classification performance. In this paper, three standard datasets, including the Indian pines dataset, Salinas-A dataset and KSC dataset, are used as experimental data to compare the application effect of the ReliefF-RFE algorithm with ReliefF and RFE algorithms. The results show that the hyperspectral image classification by applying the ReliefF-RFE algorithm has an average overall accuracy (OA) of 92.94%, F-measure of 92.81%, and Kappa coefficient of 91.94%; in the three datasets, the average feature dimension of ReliefF-RFE algorithm is 37% of that of ReliefF algorithm, while the average operation time is 75% of that of the RFE algorithm. It shows that the ReliefF-RFE algorithm can ensure the classification accuracy while overcoming the defects of the filtered ReliefF algorithm, which cannot effectively reduce the redundancy among features and the wrapped RFE algorithm, which has high time complexity and has a more balanced comprehensive performance, which is suitable for feature selection in hyperspectral image classification.
[1] DU Pei-jun, XIA Jun-shi, XUE Chao-hui, et al(杜培军,夏俊士,薛朝辉,等). National Remote Sensing Bulletin(遥感学报), 2016, 20(2): 236.
[2] REN Xiao-dong, LEI Wu-hu, GU Yu, et al(任晓东,雷武虎,谷 雨,等). Computer Science(计算机科学), 2015, 42(S2): 162.
[3] ZHANG Bing(张 兵). National Remote Sensing Bulletin(遥感学报), 2016, 20(5): 1062.
[4] LI Zhi-qin, DU Jian-qiang, NIE Bin, et al(李郅琴,杜建强,聂 斌,等). Computer Engineering and Applications(计算机工程与应用), 2019, 55(24): 10.
[5] Ren J S, Wang R X, Liu G, et al. Remote Sensing, 2020, 12(7): 1104.
[6] Ye M C, Xu C X, Chen H, et al. International Journal of Wavelets, Multiresolution and Information Processing, 2019, 17(5): 17.
[7] ZHANG Dong-yan, YANG Yu-ying, HUANG Lin-sheng, et al(张东彦,杨玉莹,黄林生,等). Transactions of the Chinese Society of Agricultural Engineering(农业工程学报), 2021, 37(9): 110.
[8] LIU Dai-chao, LI Xiao-song, LI Xiang-chen, et al(刘代超,李晓松,李向晨,等). Bulletin of Surveying and Mapping(测绘通报), 2020,(8): 5.
[9] XU Ling-ling, CHI Dong-xiang(徐玲玲,迟冬祥). Computer Engineering and Applications(计算机工程与应用), 2020, 56(24): 12.
[10] Marwa H, Slim B, Chih C H, et al. Memetic Computing, 2019, 11(2): 193.
[11] Yuan M S, Yang Z J, Huang G Z, et al. Pattern Recognition Letters, 2017, 92: 17.
[12] Lin X H, Li C, Ren W J, et al. Computational Biology and Chemistry, 2019, 83: 107149.
[13] DING Si-fan, WANG Feng, WEI Wei(丁思凡,王 锋,魏 巍). Computer Science(计算机科学), 2021, 48(4): 91.
[14] ZHANG Xiao-nei, ZHAI Wen-peng, HOU Hui-rang, et al(张小内,翟文鹏,侯惠让,等). Journal of Electronics & Information Technology(电子与信息学报), 2021, 43(7): 2032.
[15] WANG Xiang, HU Xue-gang(王 翔,胡学钢). Journal of Computer Applications(计算机应用), 2017, 37(9): 2433.