Automated Deep Learning Driven Crop Classification on Hyperspectral Remote Sensing Images

Hyperspectral remote sensing/imaging spectroscopy is a novel approach to reaching a spectrum from all the places of a huge array of spatial places so that several spectral wavelengths are utilized for making coherent images.Hyperspectral remote sensing contains acquisition of digital images from several narrow,contiguous spectral bands throughout the visible,Thermal Infrared(TIR),Near Infrared(NIR),and Mid-Infrared(MIR)regions of the electromagnetic spectrum.In order to the application of agricultural regions,remote sensing approaches are studied and executed to their benefit of continuous and quantitativemonitoring.Particularly,hyperspectral images(HSI)are considered the precise for agriculture as they can offer chemical and physical data on vegetation.With this motivation,this article presents a novel Hurricane Optimization Algorithm with Deep Transfer Learning Driven Crop Classification(HOADTL-CC)model onHyperspectralRemote Sensing Images.The presentedHOADTL-CC model focuses on the identification and categorization of crops on hyperspectral remote sensing images.To accomplish this,the presentedHOADTL-CC model involves the design ofHOAwith capsule network(CapsNet)model for generating a set of useful feature vectors.Besides,Elman neural network(ENN)model is applied to allot proper class labels into the input HSI.Finally,glowworm swarm optimization(GSO)algorithm is exploited to fine tune the ENNparameters involved in this article.The experimental result scrutiny of the HOADTL-CC method can be tested with the help of benchmark dataset and the results are assessed under distinct aspects.Extensive comparative studies stated the enhanced performance of the HOADTL-CC model over recent approaches with maximum accuracy of 99.51%.

Classification of hyperspectral remote sensing images based on simulated annealing genetic algorithm and multiple instance learning

A hybrid feature selection and classification strategy was proposed based on the simulated annealing genetic algonthrn and multiple instance learning （MIL）. The band selection method was proposed from subspace decomposition, which combines the simulated annealing algorithm with the genetic algorithm in choosing different cross-over and mutation probabilities, as well as mutation individuals. Then MIL was combined with image segmentation, clustering and support vector machine algorithms to classify hyperspectral image. The experimental results show that this proposed method can get high classification accuracy of 93.13% at small training samples and the weaknesses of the conventional methods are overcome.

STUDY ON TRANSMISSION MECHANISM OF THERMAL INFRARED REMOTE SENSING FOR ABRUPT GEOTHERMAL ANOMALY IN VOLCANIC REGION

Experiment researches have proven that there is an obvious phenomenon of abrupt geothermal anomaly in volcanic region in the forewarning period of volcano eruption, which is closely related to the geological structure, the cause, the scale and the type of volcano etc. On the other hand, this kind of geothermal anomaly is an important sign to monitor volcano activity by thermal infrared remote sensing techniques. This paper discusses the feature of abrupt geothermal anomaly, the transmission mechanism of geothermal anomaly and the radiation transmission mechanism of heat field of terrene in volcanic region. By analyzing mechanism of terrene temperature rising by way of conduction and convection of heat, we have presented the transmission equation of atmosphere for thermal infrared radiation based on the effective radiation of objects. The related problems of noise interference in the processes of transmission for thermal infrared radiation will be discussed in the later paper.

Improving wavelet reconstruction algorithm to achieve comprehensive application of thermal infrared remote sensing data from TM and MODIS

According to the data characteristics of Landsat thematic mapper （TM） and MODIS, a new fu sion algorithm about thermal infrared data has been proposed in the article based on improving wave let reconstruction. Under the domain of neighborhood wavelet reconstruction, data of TM and MO DIS are divided into three layers using wavelet decomposition. The texture information of TM data is retained by fusing highfrequency information. The neighborhood correction coefficient method （NC CM） is set up based on the search neighborhood of a certain size to fuse lowfrequency information. Thermal infrared value of MODIS data is reduced to the space value of TM data by applying NCCM. The data with high spectrum, high spatial and high temporal resolution, are obtained through the al gorithm in the paper. Verification results show that the texture information of TM data and high spec tral information of MODIS data could be preserved well by the fusion algorithm. This article could provide technical support for high precision and fast extraction of the surface environment parame ters.

INFRARED THERMAL IMAGE STUDY ON THE FOREWARNING OF COAL AND SANDSTONE FAILURE UNDER LOAD

In the experimental study, AGE-782 thermal instrument was used to detect the infrared radiation variation of coal and sandstone (wave-length range 3.6～5.5 μm was used). It's discovered that coal and sandstone failure under load have three kinds of infrared thermal features as well as infrared forewarning messages. That are: (1) temperature rises gradually but drops before failure ; (2) temperature rises gradually but quickly rises before failure; (3) first rises,then drops and lower temperature emerges before failure. The further researches and the prospect of micro-wave remote sensing detection .on ground pressure is also discussed.

Interpreting the Shortwave Infrared &Thermal Infrared Regions of Remote Sensed Electromagnetic Spectrum with Application for Mineral-Deposits Exploration

The ASTER (Advanced Space-borne Thermal Emission and Reflection radiometer) data, including all the 3 parts: VNIR (Visible and Near-Infrared), SWIR (Short Wave Infrared), TIR (Thermal Infrared), were applied for extraction of mineral deposits, such as the Ni-Cu deposit in eastern Tianshan, the gypsum in western Tianshan, and the borax in Tibetan. This paper discusses the extraction methodology using the ASTER remote sensing data and reveals the good extraction results. This paper bravely represents the summary of the main achievement for this field by the scientists in other countries and gives a comparison with the works by others. The new achievements, described in this paper, comprise the extraction of anomalies for Ni-Cu deposit, gypsum, and borax.

Monitoring coal fires in Datong coalfield using multi-source remote sensing data

Numerous coal fires burn underneath the Datong coalfield because of indiscriminate mining.Landsat TM/ETM,unmanned aerial vehicle（UAV）,and infrared thermal imager were employed to monitor underground coal fires in the Majiliang mining area.The thermal field distributions of this area in 2000,2002,2006,2007,and 2009 were obtained using Landsat TM/ETM.The changes in the distribution were then analyzed to approximate the locations of the coal fires.Through UAV imagery employed at a very high resolution（0.2 m）,the texture information,linear features,and brightness of the ground fissures in the coal fire area were determined.All these data were combined to build a knowledge model of determining fissures and were used to support underground coal fire detection.An infrared thermal imager was used to map the thermal field distribution of areas where coal fire is serious.Results were analyzed to identify the hot spot trend and the depth of the burning point.

Early detection of pine wilt disease in Pinus tabuliformis in North China using a field portable spectrometer and UAV-based hyperspectral imagery

Background:Pine wilt disease(PWD)is a major ecological concern in China that has caused severe damage to millions of Chinese pines(Pinus tabulaeformis).To control the spread of PWD,it is necessary to develop an effective approach to detect its presence in the early stage of infection.One potential solution is the use of Unmanned Airborne Vehicle(UAV)based hyperspectral images(HIs).UAV-based HIs have high spatial and spectral resolution and can gather data rapidly,potentially enabling the effective monitoring of large forests.Despite this,few studies examine the feasibility of HI data use in assessing the stage and severity of PWD infection in Chinese pine.Method:To fill this gap,we used a Random Forest(RF)algorithm to estimate the stage of PWD infection of trees sampled using UAV-based HI data and ground-based data(data directly collected from trees in the field).We compared relative accuracy of each of these data collection methods.We built our RF model using vegetation indices(VIs),red edge parameters(REPs),moisture indices(MIs),and their combination.Results:We report several key results.For ground data,the model that combined all parameters(OA:80.17%,Kappa:0.73)performed better than VIs(OA:75.21%,Kappa:0.66),REPs(OA:79.34%,Kappa:0.67),and MIs(OA:74.38%,Kappa:0.65)in predicting the PWD stage of individual pine tree infection.REPs had the highest accuracy(OA:80.33%,Kappa:0.58)in distinguishing trees at the early stage of PWD from healthy trees.UAV-based HI data yielded similar results:the model combined VIs,REPs and MIs(OA:74.38%,Kappa:0.66)exhibited the highest accuracy in estimating the PWD stage of sampled trees,and REPs performed best in distinguishing healthy trees from trees at early stage of PWD(OA:71.67%,Kappa:0.40).Conclusion:Overall,our results confirm the validity of using HI data to identify pine trees infected with PWD in its early stage,although its accuracy must be improved before widespread use is practical.We also show UAV-based data PWD classifications are less accurate but comparable to those of ground-based data.We believe that these results can be used to improve preventative measures in the control of PWD.

Research on the Estimation Model of Soil Moisture Content Based on the Characteristics of Thermal Infrared Data

With the portable Fourier Transform Infrared Spectroscopy (FTIR), the reflectance spectra of soil samples with different moisture content are measured in laboratory for expounding the characteristic of radiation in the thermal infrared part of the spectrum with different soil moisture content. A model of estimating the moisture content in soil is attempted to make based on Moisture Diagnostic Index (MDI). In general,the spectral characteristic of soil emissivity in laboratory includes the following aspects.Firstly,in the region of 8.0-9.5 μm,along with the increase of soil moisture content,the emissivity of soil increases to varying degrees. The spectral curves are parallel relatively and have a tendency to become horizontal and the absorbed characteristic of reststrahlen is also weakened relatively with the increase of soil moisture in this region.Secondly,in the region of 11.0-14.0 μm,the emissivity of soil has a tendency of increasing.There is an absorption value near about 12.7 μm. As the soil moisture content increases,the depth of absorption also increases. This phenomenon may be caused by soil moisture absorption. Methods as derivative, difference and standardized ratio transformation may weaken the background noise effectively to the spectrum data. Especially using the ratio of the emissivity to the average of 8-14 μm may obviously enhance the correlation between soil moisture and soil emissivity. According to the result of correlation analysis, the 8.237 μm is regarded as the best detecting band for soil moisture content. Moreover,based on the Moisture Diagnostic Index ( MDI) in the 8.194-8.279 μm, the logarithmic model of estimating soil moisture is made.

Impacts on Initial Condition Modification from Hyperspectral Infrared Sounding Data Assimilation: Comparisons between Full-Spectrum and Channel-Selection Scheme Based on Two-Month Experiments Using CrIS and IASI Observation

This paper discusses the performance difference between full-spectrum and channel-selection assimilation scheme of hyperspectral infrared observation, e.g. CrIS and IASI, on improving the accuracy of initial condition in numerical weather prediction. To accomplish this, we develop a 3D-Variational data assimilation system whose observation operator is a principal-component based fast radiative transfer model, which equips the direct assimilation of full-channel radiance from hyperspectral infrared sounders with high computational efficiency. This project’s primary goal is to demonstrate that assimilation of infrared observation in a full-channel mode could improve the accuracy of initial condition compared to selected-channel assimilation. Results show that full-channel assimilation performs better than selected-channel assimilation in modifying low and middle troposphere (1000 - 700 hPa, 700 - 400 hPa) temperature and water vapor field, while marginal improvements from temperature and water vapor field could be found over upper troposphere (400 - 100 hPa). This research also proves the feasibility of an alternative path to data assimilation for the full usage of hyperspectral infrared sounding observation in numerical weather prediction.

Formation cause of thermal infrared high temperature belt along Honghe fault and its relation to earthquakes

Aiming at two Dayao earthquakes with magnitude more than 6 occurred in 2003 in Yunnan Province, we analyzed and interpreted the NOAA satellite thermal infrared images of 1999, 2003 and 2004 in Chuandian region, and also calculated the annual variation of brightness temperature of the hot belt along Honghe fault to explore the formation cause of the high temperature belt and its relation to the earthquakes. The results show that the high temperature belt along Honghe fault is caused by geographic environment factors, such as water system and terrain. But the annual average brightness temperature of the belt in earthquake year of 2003 is clearly higher than that in no earthquake years of 1999 and 2004, this maybe indicates that the thermal activities of Honghe fault increase in earthquake years, and can cause the annual variation anomaly of brightness temperature. We can detect and monitor this thermal activities of Honghe fault before earthquake by analyzing and comparing the relative changes of thermal infrared brightness temperature of the hot belt in different years.

Retrieval of urban land surface component temperature using multi-source remote-sensing data

The components of urban surface cover are diversified,and component temperature has greater physical significance and application values in the studies on urban thermal environment.Although the multi-angle retrieval algorithm of component temperature has been matured gradually,its application in the studies on urban thermal environment is restricted due to the difficulty in acquiring urban-scale multi-angle thermal infrared data.Therefore,based on the existing multi-source multi-band remote sensing data,access to appropriate urban-scale component temperature is an urgent issue to be solved in current studies on urban thermal infrared remote sensing.Then,a retrieval algorithm of urban component temperature by multi-source multi-band remote sensing data on the basis of MODIS and Landsat TM images was proposed with expectations achieved in this work,which was finally validated by the experiment on urban images of Changsha,China.The results show that：1） Mean temperatures of impervious surface components and vegetation components are the maximum and minimum,respectively,which are in accordance with the distribution laws of actual surface temperature; 2） High-accuracy retrieval results are obtained in vegetation component temperature.Moreover,through a contrast between retrieval results and measured data,it is found that the retrieval temperature of impervious surface component has the maximum deviation from measured temperature and its deviation is greater than 1 ℃,while the deviation in vegetation component temperature is relatively low at 0.5 ℃.

FRAME DESIGN OF REMOTE SENSING MONITORINGFOR VOLCANIC ACTIVITIES IN CHANGBAI MOUNTAINS

Volcanic eruption is one of the most serious geological disasters,however,a host of facts have proven that the Changbai Mountains volcano is a modern dormant one and has ever erupted disastrously. With the rapid development of remote sensing technology,space monitoring of volcanic activities has already become possible,particularly in the application of thermal infrared remote sensing. The paper,through the detailed analysis of geothermal anomaly factors such as heat radiation,heat conduction and convection,depicts the monitoring principles by which volcano activities would be monitored efficiently and effectively. Reasons for abrupt geothermal anomaly are mainly analyzed,and transmission mechanism of geothermal anomaly in the volcanic regions is explained. Also,a variety of noises disturbing the transmission of normal geothermal anomaly are presented. Finally,some clues are given based on discussing thermal infrared remote sensing monitoring mechanism toward the volcanic areas.

Analysis of Thermal Infrared Anomalies for the April16,2013M_W7.8of Khash, Iran Earthquake

After using the "Time-Frequency Relative Power Spectrum"( T-F RPS) method based on the China Geostationary Meteorological Satellite( FY-2 C/FY-2 E) infrared remote sensing brightness temperature data processing,we rapidly and accurately extracted and identified pre-earthquake thermal infrared anomalies for the April 16,2013 MW7. 8 of Khash,Iran Earthquake. Spatial evolution of anomalies showed the distribution and process. The anomalies were mainly distributed in the east of Khash,Iran. The characteristics of process and distribution presented X-Type model of NE and near NS strip which relates to the geological structure of this region. The epicenter was located near the intersection region of the X-Type abnormal migration process. Besides,the results of time series of anomalies showed that,the duration was more than 40 days and the maximum amplitude was about18 times. The earthquake occurred 20 days after the abnormal maximum amplitude which appeared on March 26,2013.

Mapping of Aluminous Rich Laterite Depositions through Hyper Spectral Remote Sensing

Increased dimensionality of the satellite data proves to be very useful for discriminating features with very close spectral matching. Present study concentrates on the retrieval of reflectance spectra from the level one radiometrically corrected data in Koraput district (Orissa) for the Bauxite ore. In the present study, atmospheric correction model FLAASH has been used to retrieve reflectance from the radiance data. Preprocessing of the dataset has been done before applying atmospheric correction on the dataset. Spectral subsetting of noise prone bands has been successfully done. Local destriping of the affected bands has been done using a 3*3 local mean filter. Spectral signatures of samples were derived from the processed data. Spectral signature of each sample and derived features vectors were correlated with the satellite image of the area and distribution of each feature was demarcated. Spatial abundance of each feature was used in preparation of mineral abundance map. Accuracy of the map was assessed using training sets of representative geological units. The mineral abundance mapping using the spectral analysis of the reflectance image involves the endmember collection using the N-Dimensional visualizer tool in ENVI software. Laterite, Bauxite, Iron and silica rich Aluminous laterite soil, Alluvium and Forest were selected as the end members after understanding the geology and analysis of the reflectance image. Various mapping techniques were applied to generate the final classified mineral abundance Map, Linear Spectral Unmixing, Mixture Tune Matched Filtering, Spectral Feature Fitting, Spectral Angle Mapper were the techniques used. Results have revealed the ability of Hyper spectral Remote sensing data for the identification and mapping of Hydrothermal altered products like Bauxite, Aluminous Laterite. This technology can be utilized for targeting minerals in the altered zone.

基于改进2DCNN的高光谱遥感图像处理研究

针对传统遥感图像处理中的时间成本和人工成本高、效率低等问题,以提高遥感高光谱图像分类中的处理速度、精度,降低参数量为目标,提出改进的2DCNN模型En-De-2CP-2DCNN.首先,使用1DCNN、2DCNN与3DCNN在Pavia University HSI数据集上分别进行分类实验,对比分析各自优缺点.其次,在保持较快的处理速度和不增加模型参数量的前提下,选择2DCNN为基础模型,参考SegNet的Encoder-Decoder结构,融入双卷积池化思想进行基础模型改进,同时优化学习策略.结果表明:En-De-2CP-2DCNN模型F1为99.96%,达到3DCNN的同等水平(99.36%),较改进前(97.28%)提高2.68个百分点;处理速度(5 s/epoch)和1DCNN位于同一量级,快于3DCNN(96 s/epoch);参数量(2.01 MB)较改进前降低了1.54 MB,虽高于3DCNN(316 KB),但远低于1DCNN(19.21 MB).En-De-2CP-2DCNN模型在处理速度和参数量方面的改进,有利于进一步实现移动端的轻量化部署.

IspecHyper多旋翼无人机高光谱影像处理方法研究

为解决IspecHyper(莱森光学)多旋翼无人机高光谱成像系统缺乏数据处理配套软件,采集多航带高光谱数据误差大、坐标缺失、无法自动拼接等问题,以武鸣区太平镇角龙村柑橘种植基地为研究区,开展IspecHyper多旋翼无人机高光谱影像处理方法研究。首先,利用IspecHyper-VM200成像系统获取研究区高清照片和多航带高光谱影像数据;其次,以高清照片为数据源,通过PXI4D Mapper软件预处理和ENVI软件影像几何校正,形成高分辨率无人机正射影像;最后,利用ENVI软件裁剪多航带高光谱影像扭曲边界数据,以无人机正射影像为基准完成几何校正,进而通过影像镶嵌和光谱转换计算,形成高光谱反射率影像产品。结果表明,该研究形成的技术方法可有效解决IspecHyper多旋翼无人机高光谱影像处理存在问题,同时为无人机高光谱影像处理提供技术参考。

基于多位移光谱遥感图像的空间引力模型亚像元定位

目前基于多位移光谱遥感图像的亚像元定位方法通常很少考虑点扩散函数效应影响,本文提出了一种基于多位移光谱遥感图像的空间引力模型亚像元定位方法。为了生成粗糙丰度图像,首先对多个多位移光谱遥感图像进行解混;在点扩散函数效应被考虑的前提下,对粗糙丰度图像实施面积到点的克里插值处理,然后通过理想方波滤波进行滤波,最终获得改进后的粗糙丰度图像;利用空间引力模型对改进后的粗糙丰度图像进行上采样,从而得到上采样丰度图像,再对上采样丰度图像执行整合处理,以生成精细丰度图像;在完成上述所有图像处理步骤后,最终通过应用类别分配方法,将类别标签分发给各个亚像元,以此获得精确的定位结果。在2组实验数据集上的实验结果表明:本文提出的方法比现有的亚像元定位方法获得了更好的定位效果。

基于ADMM和深度生成先验的高光谱解混方法

混合像元的存在制约了高光谱图像分类和目标检测的精度,为了提高混合像元分解的精度,精确地分析混合像元中的组成成分,本文提出将优化方法和深度生成先验结合的高光谱解混方法,实现数据驱动和模型驱动的有机结合。近年来,基于深度神经网络的处理方法被广泛使用在高光谱解混任务中。但是该类方法是“黑盒子”,缺乏物理可解释性。传统的基于数学优化的高光谱解混方法,通过使用人工设计的先验项引入图像内含信息,提高解混精度。但是对于复杂的先验项,求解方法复杂,且并不是所有先验信息都可以用数学模型表示出来。所以本文提出一结合交替方向乘子法优化算法和深度生成先验的高光谱解混方法,联合使用数学优化和深度方法的优越性。首先使用ADMM算法将数据拟合项和生成先验项进行解耦,对于生成先验,使用传统解混方法获得的丰度预训练变分自编码器网络,并将解码器作为生成器。本文同时使用人工合成数据和真实遥感数据验证所提出方法的有效性。

基于多时相热红外遥感的理塘盆地地热资源探测

地热能是地球内部一种环保、可再生的地下热能,热量传导至地表会引起地表温度异常。然而,地表温度受到太阳辐射、周期性变化和降水等诸多因素的影响,可能掩盖由地热引起的温度异常。为了解决这些问题并突出地热引起的地表温度异常,文中提出了一种利用Google Earth Engine(GEE)高效快速计算多时相冬季平均地表温度并提取地热异常的方法,并对理塘盆地的地热资源分布特征进行了研究。该方法综合运用多时相冬季平均地表温度对比分析、水体剔除、地形效应抑制等手段来提取地热异常区。结果表明,文中确定的8处地热异常区与断裂构造关系密切,且区内水资源丰富,具备地热资源形成的良好条件,可以作为未来地热勘探的重点方向。研究认为,在高原恶劣环境中,利用多时相平均冬季地表温度识别地热异常是一种高效且经济的方法。