《计算机应用》唯一官方网站 ›› 2022, Vol. 42 ›› Issue (2): 485-492.DOI: 10.11772/j.issn.1001-9081.2021020332
收稿日期:
2021-03-05
修回日期:
2021-04-30
接受日期:
2021-05-07
发布日期:
2022-02-11
出版日期:
2022-02-10
通讯作者:
陈红梅
作者简介:
翟东昌(1993—),男,甘肃兰州人,硕士研究生,主要研究方向:数据挖掘;基金资助:
Dongchang ZHAI1, Hongmei CHEN2()
Received:
2021-03-05
Revised:
2021-04-30
Accepted:
2021-05-07
Online:
2022-02-11
Published:
2022-02-10
Contact:
Hongmei CHEN
About author:
ZHAI Dongchang, born in 1993, M. S. candidate. His research interests include data mining.Supported by:
摘要:
为了减少高光谱图像数据中的冗余信息,优化计算效率,并提升图像数据后续应用的有效性,提出一种基于邻域熵(NE)的高光谱波段选择算法。首先,为了高效计算样本的邻域子集,采用了局部敏感哈希(LSH)作为近似最近邻的搜索策略;然后,引入了NE理论来度量波段和类之间的互信息(MI),并把最小化特征集合与类变量之间的条件熵作为选取有效波段的方法;最后,采用两个数据集,通过支持向量机(SVM)和随机森林(RM)进行分类实验。实验结果表明,相较于四种基于MI的特征选择算法,从总体精度以及Kappa系数上看,所提算法能够在30个波段内较快地选取有效波段子集,并达到局部最优。该算法的部分实验结果的总体精度以及Kappa系数分别达到全局最优的92.99%以及0.860 8,表明所提算法能有效地处理高光谱波段选择问题。
中图分类号:
翟东昌, 陈红梅. 基于邻域熵的高光谱波段选择算法[J]. 计算机应用, 2022, 42(2): 485-492.
Dongchang ZHAI, Hongmei CHEN. Hyperspectral band selection algorithm based on neighborhood entropy[J]. Journal of Computer Applications, 2022, 42(2): 485-492.
类标 | 类名 | 样本数 |
---|---|---|
1 | Alfalfa | 46 |
2 | Corn-notill | 1 428 |
3 | Corn-mintill | 830 |
4 | Corn | 237 |
5 | Grass-pasture | 483 |
6 | Grass-trees | 730 |
7 | Grass-pasture-mowed | 28 |
8 | Hay-windrowed | 478 |
9 | Oats | 20 |
10 | Soybean-notill | 972 |
11 | Soybean-mintill | 2 455 |
12 | Soybean-clean | 593 |
13 | Wheat | 205 |
14 | Woods | 1 265 |
15 | Buildings-Grass-Trees-Drives | 386 |
16 | Stone-Steel-Towers | 93 |
表1 Indian Pines数据集的类标以及相应的样本数
Tab. 1 Class labels and corresponding numbers of samples in Indian Pines dataset
类标 | 类名 | 样本数 |
---|---|---|
1 | Alfalfa | 46 |
2 | Corn-notill | 1 428 |
3 | Corn-mintill | 830 |
4 | Corn | 237 |
5 | Grass-pasture | 483 |
6 | Grass-trees | 730 |
7 | Grass-pasture-mowed | 28 |
8 | Hay-windrowed | 478 |
9 | Oats | 20 |
10 | Soybean-notill | 972 |
11 | Soybean-mintill | 2 455 |
12 | Soybean-clean | 593 |
13 | Wheat | 205 |
14 | Woods | 1 265 |
15 | Buildings-Grass-Trees-Drives | 386 |
16 | Stone-Steel-Towers | 93 |
类标 | 类名 | 样本数 |
---|---|---|
1 | Brocoli_green_weeds_1 | 2 009 |
2 | Brocoli_green_weeds_2 | 3 726 |
3 | Fallow | 1 976 |
4 | Fallow_rough_plow | 1 394 |
5 | Fallow_smooth | 2 678 |
6 | Stubble | 3 959 |
7 | Celery | 3 579 |
8 | Grapes_untrain | 11 271 |
9 | Soil_vinyard_develop | 6 203 |
10 | Corn_senesced_green_weeds | 3 278 |
11 | Lettuce_romaine_4wk | 1 068 |
12 | Lettuce_romaine_5wk | 1 927 |
13 | Lettuce_romaine_6wk | 916 |
14 | Lettuce_romaine_7wk | 1 070 |
15 | Vinyard_untrain | 7 268 |
16 | Vinyard_vertical_trellis | 1 807 |
表2 Salinas数据集的类标以及相应的样本数
Tab. 2 Class labels and corresponding numbers of samples in Salinas dataset
类标 | 类名 | 样本数 |
---|---|---|
1 | Brocoli_green_weeds_1 | 2 009 |
2 | Brocoli_green_weeds_2 | 3 726 |
3 | Fallow | 1 976 |
4 | Fallow_rough_plow | 1 394 |
5 | Fallow_smooth | 2 678 |
6 | Stubble | 3 959 |
7 | Celery | 3 579 |
8 | Grapes_untrain | 11 271 |
9 | Soil_vinyard_develop | 6 203 |
10 | Corn_senesced_green_weeds | 3 278 |
11 | Lettuce_romaine_4wk | 1 068 |
12 | Lettuce_romaine_5wk | 1 927 |
13 | Lettuce_romaine_6wk | 916 |
14 | Lettuce_romaine_7wk | 1 070 |
15 | Vinyard_untrain | 7 268 |
16 | Vinyard_vertical_trellis | 1 807 |
算法 | 时间复杂度 |
---|---|
NEFS/SINEFS | O(DNαs2) |
mRMR-MIFS | O(DNs2) |
X-MIFS | O(DNs2) |
NMI | O(DNs log N) |
WaLuMI | O(LDN+L2D) |
表3 各个算法的时间复杂度
Tab. 3 Time complexity of each algorithm
算法 | 时间复杂度 |
---|---|
NEFS/SINEFS | O(DNαs2) |
mRMR-MIFS | O(DNs2) |
X-MIFS | O(DNs2) |
NMI | O(DNs log N) |
WaLuMI | O(LDN+L2D) |
算法 | 系数 | 截距 |
---|---|---|
SINEFS | 1.02×10-5 | 16.600 |
NEFS | 2.61×10-4 | 257.000 |
mRMR-MIFS | 1.57×10-6 | -0.152 |
X-MIFS | 4.8×10-7 | -0.144 |
NMI | 3.8×10-5 | -0.521 |
WaLuMI | 3.2×10-6 | -0.489 |
表4 CPU时间线性模型
Tab. 4 CPU time linear model
算法 | 系数 | 截距 |
---|---|---|
SINEFS | 1.02×10-5 | 16.600 |
NEFS | 2.61×10-4 | 257.000 |
mRMR-MIFS | 1.57×10-6 | -0.152 |
X-MIFS | 4.8×10-7 | -0.144 |
NMI | 3.8×10-5 | -0.521 |
WaLuMI | 3.2×10-6 | -0.489 |
1 | AMIGO J M, BABAMORADI H, ELCOROARISTIZABAL S. Hyperspectral image analysis. a tutorial[J]. Analytica Chimica Acta, 2015, 896: 34-51. 10.1016/j.aca.2015.09.030 |
2 | GU Y F, LIU T Z, GAO G M, et al. Multimodal hyperspectral remote sensing: an overview and perspective[J]. Science China — Information Sciences, 2021, 64(2): No.121301. 10.1007/s11432-020-3084-1 |
3 | JANSE P V, DESHMUKH R R. Hyperspectral remote sensing for agriculture: a review[J]. International Journal of Computer Applications, 2017, 172(7): 30-34. |
4 | TURPIE K, VERAVERBEKE S, WRIGHT R, et al. NASA 2014 the Hyperspectral Infrared Imager (HyspIRI) — science impact of deploying instruments on separate platforms: JPL Publication 14-13[R]. Pasadena, CA: Jet Propulsion Laboratory of California Institute of Technology, 2014. |
5 | HUGHES G. On the mean accuracy of statistical pattern recognizers[J]. IEEE Transactions on Information Theory, 1968, 14(1): 55-63. 10.1109/tit.1968.1054102 |
6 | BELLMAN R. Adaptive Control Processes: a Guided Tour[M]. Princeton: Princeton University Press, 2015: 3-257. |
7 | SUN W W, HALEVY A, BENEDETTO J J, et al. UL-Isomap based nonlinear dimensionality reduction for hyperspectral imagery classification[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2014, 89: 25-36. 10.1016/j.isprsjprs.2013.12.003 |
8 | UDDIN M P, MAMUN M A, HOSSAIN M A. PCA-based feature reduction for hyperspectral remote sensing image classification[J]. IETE Technical Review, 2021, 38(4): 377-396. 10.1080/02564602.2020.1740615 |
9 | SUN W W, DU Q. Hyperspectral band selection: a review[J]. IEEE Geoscience and Remote Sensing Magazine, 2019, 7(2): 118-139. 10.1109/mgrs.2019.2911100 |
10 | 张伍,陈红梅.基于核模糊粗糙集的高光谱波段选择算法[J].计算机应用, 2019, 40(1): 258-263. |
ZHANG W, CHEN H M. Hyperspectral band selection algorithm based on kernelized fuzzy rough set[J]. Journal of Computer Applications, 2019, 40(1): 258-263. | |
11 | MOSLEHI F, HAERI A. A novel hybrid wrapper-filter approach based on genetic algorithm, particle swarm optimization for feature subset selection[J]. Journal of Ambient Intelligence and Humanized Computing, 2020, 11(3): 1105-1127. 10.1007/s12652-019-01364-5 |
12 | ZHANG L F, ZHANG Q, ZHANG L P, et al. Ensemble manifold regularized sparse low-rank approximation for multiview feature embedding[J]. Pattern Recognition, 2015, 48(10): 3102-3112. 10.1016/j.patcog.2014.12.016 |
13 | CHANDRASHEKAR G, SAHIN F. A survey on feature selection methods[J]. Computers and Electrical Engineering, 2014, 40(1): 16-28. 10.1016/j.compeleceng.2013.11.024 |
14 | HU L, GAO W, ZHAO K, et al. Feature selection considering two types of feature relevancy and feature interdependency[J]. Expert Systems with Applications, 2018, 93: 423-434. 10.1016/j.eswa.2017.10.016 |
15 | TAO H, HOU C P, NIE F P, et al. Effective discriminative feature selection with nontrivial solution[J]. IEEE Transactions on Neural Networks and Learning Systems, 2016, 27(4): 796-808. 10.1109/tnnls.2015.2424721 |
16 | KIRA K, RENDELL L A. A practical approach to feature selection [C]// Proceedings of the 9th International Workshop on Machine Learning. San Francisco: Morgan Kaufmann Publishers Inc., 1992: 249-256. 10.1016/b978-1-55860-247-2.50037-1 |
17 | SHARMIN S, ALI A A, KHAN M A H, et al. Feature selection and discretization based on mutual information [C]// Proceedings of the 2017 IEEE International Conference on Imaging, Vision and Pattern Recognition. Piscataway: IEEE, 2017: 1-6. 10.1109/icivpr.2017.7890885 |
18 | GUO B, GUNN S R, DAMPER R I, et al. Band selection for hyperspectral image classification using mutual information[J]. IEEE Geoscience and Remote Sensing Letters, 2006, 3(4): 522-526. 10.1109/lgrs.2006.878240 |
19 | AMANKWAH A. Spatial mutual information based hyperspectral band selection for classification[J]. The Scientific World Journal, 2015, 2015: No.630918. 10.1155/2015/630918 |
20 | LIU Y, XIE H, CHEN Y H, et al. Neighborhood mutual information and its application on hyperspectral band selection for classification[J]. Chemometrics and Intelligent Laboratory Systems, 2016, 157: 140-151. 10.1016/j.chemolab.2016.07.009 |
21 | FANO R M. Transmission of information: a statistical theory of communications[J]. American Journal of Physics, 1961, 29(11): 793-794. 10.1119/1.1937609 |
22 | JIANG Y, LI C Y. mRMR-based feature selection for classification of cotton foreign matter using hyperspectral imaging[J]. Computers and Electronics in Agriculture, 2015, 119: 191-200. 10.1016/j.compag.2015.10.017 |
23 | KVÅLSETH T O. On normalized mutual information: measure derivations and properties[J]. Entropy, 2017, 19(11): No.631. 10.3390/e19110631 |
24 | BRUNATO M, BATTITI R. X-MIFS: exact mutual information for feature selection [C]// Proceedings of the 2016 International Joint Conference on Neural Networks. Piscataway: IEEE, 2016: 3469-3476. 10.1109/ijcnn.2016.7727644 |
25 | MARTÍNEZ-USÓ A, PLA F, SOTOCA J M, et al. Clustering-based hyperspectral band selection using information measures[J]. IEEE Transactions on Geoscience and Remote Sensing, 2007, 45(12): 4158-4171. 10.1109/tgrs.2007.904951 |
26 | SUN L, ZHANG X Y, QIAN Y H, et al. Feature selection using neighborhood entropy-based uncertainty measures for gene expression data classification[J]. Information Sciences, 2019, 502: 18-41. 10.1016/j.ins.2019.05.072 |
27 | GIONIS A, INDYK P, MOTWANI R. Similarity search in high dimensions via hashing [C]// Proceedings of the 25th International Conference on Very Large Data Base. San Francisco: Morgan Kaufmann Publishers Inc., 1999: 518-529. 10.1109/icde.2000.839448 |
[1] | 李顺勇, 李师毅, 胥瑞, 赵兴旺. 基于自注意力融合的不完整多视图聚类算法[J]. 《计算机应用》唯一官方网站, 2024, 44(9): 2696-2703. |
[2] | 陈都, 李圆媛, 陈彧. 基于t检验和逐步网络搜索的有向基因调控网络推断算法[J]. 《计算机应用》唯一官方网站, 2024, 44(1): 199-205. |
[3] | 李瀚臣, 张顺香, 朱广丽, 王腾科. 基于拼音相似度的中文谐音新词发现方法[J]. 《计算机应用》唯一官方网站, 2023, 43(9): 2715-2720. |
[4] | 劳景欢, 黄栋, 王昌栋, 赖剑煌. 基于视图互信息加权的多视图集成聚类算法[J]. 《计算机应用》唯一官方网站, 2023, 43(6): 1713-1718. |
[5] | 孙林, 黄金旭, 徐久成. 基于邻域容差互信息和鲸鱼优化算法的非平衡数据特征选择[J]. 《计算机应用》唯一官方网站, 2023, 43(6): 1842-1854. |
[6] | 夏进, 王正群, 朱世明. 基于时间序列分解的交通流量预测模型[J]. 《计算机应用》唯一官方网站, 2023, 43(4): 1129-1135. |
[7] | 孙林, 马天娇, 薛占熬. 基于Fisher score与模糊邻域熵的多标记特征选择算法[J]. 《计算机应用》唯一官方网站, 2023, 43(12): 3779-3789. |
[8] | 赖星锦, 郑致远, 杜晓颜, 徐莎, 杨晓君. 基于超像素锚图二重降维的高光谱聚类算法[J]. 《计算机应用》唯一官方网站, 2022, 42(7): 2088-2093. |
[9] | 杨世刚, 刘勇国. 融合语料库特征与图注意力网络的短文本分类方法[J]. 《计算机应用》唯一官方网站, 2022, 42(5): 1324-1329. |
[10] | 陈永波, 李巧勤, 刘勇国. 基于动态相关性的特征选择算法[J]. 《计算机应用》唯一官方网站, 2022, 42(1): 109-114. |
[11] | 张伍, 陈红梅. 基于多核模糊粗糙集与蝗虫优化算法的高光谱波段选择[J]. 计算机应用, 2020, 40(5): 1425-1430. |
[12] | 张凯琳, 阎庆, 夏懿, 章军, 丁云. 基于焦点损失的半监督高光谱图像分类[J]. 计算机应用, 2020, 40(4): 1030-1037. |
[13] | 程玉胜, 宋帆, 王一宾, 钱坤. 基于专家特征的条件互信息多标记特征选择算法[J]. 《计算机应用》唯一官方网站, 2020, 40(2): 503-509. |
[14] | 曾梦, 宁彬, 蔡之华, 谷琼. 使用深度对抗子空间聚类实现高光谱波段选择[J]. 《计算机应用》唯一官方网站, 2020, 40(2): 381-385. |
[15] | 王煜, 徐建民. 用于网络新闻热点识别的热点新词发现[J]. 计算机应用, 2020, 40(12): 3513-3519. |
阅读次数 | ||||||
全文 |
|
|||||
摘要 |
|
|||||