.png){kind=icon}
返回
A Novel Grey Wolf Optimisation based CNN Classifier for Hyperspectral Image classification
.png){kind=icon}
PRE
AI
DOI：10.1007/s11042-022-12628-2.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
Hyperspectral image (HI) analysis is becoming popular in remote sensing applications due to its high spectral resolution along with high spatial resolution compared to a Multispectral image. Classification of pixel vectors in a hyperspectral image (HI) to their respective classes is challenging. HI classification is an elementary task for land cover mapping, mineral exploitation, precision agriculture, etc. Optimal parameters are required to reduce the losses in the Convolutional Neural Networks (CNNs) and provide the most accurate results possible. Studies in this regard, so far have been made with manual selection of optimal parameters using traditional trial-and-error methods like the selection of loss function, a number of convolution filters, optimizer function, etc., and found to be strenuous and time-consuming. To alleviate these challenges of selecting the hyperparameters and observing the accuracy until a competitive value is reached, this paper uses a novel mechanism to classify Hyperspectral Images using Convolutional Neural Network (CNN) where the 6-hyperparameters of CNN are optimized with Grey Wolf Optimizer (GWO). The proposed GWO-based-CNN-HI model exhibits better classification accuracy (99.95%, 99.96%, and 99.99%) on three benchmark HI datasets in comparison to traditional models. Thus the novel GWO-based-CNN-HI model finds its suitability in applications of land cover classification, crop stage detection, specially in remote applications with limited computing power.
Keyword:
Feature extraction
Stationary Wavelet Transform (SWT)
Principal Component Analysis (PCA)
Hyperspectral Image classification
Deep Learning (DL)
Convolutional Neural Network (CNN)
Grey Wolf Optimizer (GWO)
Optimal Hyper-parameters
期刊
IF:
3
论文数:
2.0W
被引数:
3.2W
机构
引用论文
On the Development and Applications of Cellulosic Nanofibrillar and Nanocrystalline Materials纤维素纳米原纤和纳米晶材料的发展与应用
PREAI
分享收藏
525: CHARACTERISTICS AND COMPLICATIONS OF TUMOR LYSIS SYNDROME IN HEMATOLOGIC MALIGNANCIES525: 肿瘤溶解综合征在血液系统恶性肿瘤中的特征与并发症
PREAI
分享收藏分享收藏Complete Azoospermia Factor b Deletion of Y Chromosome in an Infertile Male With Severe Oligoasthenozoospermia: Case Report and Literature Review.png)
Urology
UrologyIF0
PREAI
分享收藏
查看更多内容