A survey of remote sensing image classification based on CNNs

With the development of earth observation technologies,the acquired remote sensing images are increasing dramatically,and a new era of big data in remote sensing is coming.How to effectively mine these massive volumes of remote sensing data are new challenges.Deep learning provides a new approach for analyzing these remote sensing data.As one of the deep learning models,convolutional neural networks(CNNs)can directly extract features from massive amounts of imagery data and is good at exploiting semantic features of imagery data.CNNs have achieved remarkable success in computer vision.In recent years,quite a few researchers have studied remote sensing image classification using CNNs,and CNNs can be applied to realize rapid,economical and accurate analysis and feature extraction from remote sensing data.This paper aims to provide a survey of the current state-of-the-art application of CNN-based deep learning in remote sensing image classification.We first briefly introduce the principles and characteristics of CNNs.We then survey developments and structural improvements on CNN models that make CNNs more suitable for remote sensing image classification,available datasets for remote sensing image classification,and data augmentation techniques.Then,three typical CNN application cases in remote sensing image classification:scene classification,object detection and object segmentation are presented.We also discuss the problems and challenges of CNN-based remote sensing image classification,and propose corresponding measures and suggestions.We hope that the survey can facilitate the advancement of remote sensing image classification research and help remote-sensing scientists to tackle classification tasks with the state-of-art deep learning algorithms and techniques.

Low threshold random lasing in dye-doped and strongly disordered chiral liquid crystals

Random lasing was experimentally investigated in pyrromethene 597-doped strongly disordered chiral liquid crystals (CLCs) composed of the nematic liquid crystal SLC1717 and the chiral agent CB15. The concentration of the chiral agent tuned the bandgap, and disordered CLC microdomains were achieved by fast quenching of the mixture from the isotropic to the cholesteric phase. Random lasing and band edge lasing were observed synchronously, and their behavior changed with the spectral location of the bandgap. The emission band for band edge lasing shifted with the change of the bandgap, while the emission band for random lasing remained practically constant. The results show that the threshold for random lasing sharply decreases if the CLC selective reflection band overlaps with the fluorescence peak of the dye molecules and if the band edge coincides at the same time with the excitation wavelength.

Simultaneous Strain and Temperature Measurement Based on Chaotic Brillouin Optical Correlation-Domain Analysis in Large-Effective-Area Fibers

Chaotic Brillouin optical correlation domain analysis(BOCDA)has been proposed and experimentally demonstrated with the advantage of high spatial resolution.However,it faces the same issue of the temperature and strain cross-sensitivity.In this paper,the simultaneous measurement of temperature and strain can be preliminarily achieved by analyzing the two Brillouin frequencies of the chaotic laser in a large-effective-area fiber(LEAF).A temperature resolution of 1℃ and a strain resolution of 20μξ can be obtained with a spatial resolution of 3.9cm.The actual temperature and strain measurement errors are 0.37℃ and 10μξ,respectively,which are within the maximum measurement errors.