DE-striping hype spectral Remote Sensing Images using Deep Convolutional Neural Network

Hyperspectral far off detecting records reflectance or emittance information in a huge amount of bordering and tight unearthly groups, and accordingly has numerous data in distinguishing and planning the mineral zones. Then again, the science and natural information gives us some other productive data about the actual qualities of pictures and channels that have been recorded from the surface. In this work, we focus on de-striping the hyperspectral remote sensing images on Hyperion data by applying Deep Convolutional Neural Network (DCNN). What is clear is the high significance of applying the sufficient pre-preparing on Hyperion information as a result of low sign to-commotion proportion. By contrasting the known layers of DCNN model for de-striping hyperspectral pictures. The results obtained by applying the mentioned methods, it is revealed that all the higher stripes in an image as well as black color has been reduced and entirely associated with the Hyperion data alteration, and in contrast, the Hyperion imagery successfully corresponds to the de-striping of hyperspectral image with an accuracy of 91.89% using DCNN model. The proposed DCNN is capable of reaching high accuracy 150s after the start of the evaluation phase and never reaches low accuracy. The pre-trained DCNN model approach would be an adequate solution considering de-striping as its high inference time is lower compared existing available methods which are not as efficient for de-striping.

665 Spatial single-cell analysis of colorectal cancer tumour using multiplexed imaging mass cytometry

BackgroundCancer research experiments often require the dissociation of cells from their native tissue before molecular profiling, leading to the loss of spatial tissue context. The cancer genomics research has shifted from mostly profiling tumour DNA mutations towards the current frontier of investigating individual genes and gene products in single cells and their immediate microenvironments. Information at this level with the spatial context enables us to link cancer–causing mutations and environmental factors to outcomes in cell signalling, responses and survival that will lead to solutions for diagnosing, predicting progression and treating cancers in different individuals. In this project we aim to capture tissue morphology, cancer cell types, multi-parameter protein contents of single cells in within morphologically intact tissue sections of colorectal tumours from 52 patients.MethodsUsing Hyperion Imaging Mass Cytometry (IMC), we simultaneously profiled 16 protein markers for each tissue section, capturing molecular signatures of tissue architecture, cancer cells, and immune cells. IMC uses laser beam to accurately ablate every 1µm2 of tissue region, generating data at subcellular resolution for FFPE tissue sections on a glass microscopy slide. We selected 2–8 regions of interest (ROI), each containing approximately 2098 cells. The ROI sizes range from 141µm x 500µm to 1121µm x 1309µm. We developed an analysis pipeline to process raw Hyperion imaging data (IMCtools), define cellular masks with information about nuclei, membrane, cytoplasm (using CellProfiler and Ilastick), and analyses cellular communities (HistoCAT). We also generated whole exome sequencing data and histopathological mages from sections of the same tissue blocks.ResultsBy measuring 16 multiplexed proteins, for each tissue region we were able to identify up to seven cell types and preserved their spatial location within the tissue (figure 1A). Through the spatial map of the cell types to the tissue, we showed the heterogeneity of the tumour microenvironment, such as the infiltration of macrophages and B-cells to the cancer regions (figure 1A). We found cancer cells consistently marked as positive for p53 and Ki67 proteins. Moreover, we could measure the level of p53 in every individual cell within each tissue section (figure 1B). The quantitative measurement of p53 by imaging mass cytometry was correlated with the result from traditional genomic sequencing of p53 mutations and with the histopathological annotation.Abstract 665 Figure 1Characterizing the complexity of colorectal cancer. A) Cell types within a region of interest, defined by 16 markers. Cancer cells are consistent to the. B) Quantifying p53 expression from Hyperion dataConclusionsApplying the Hyperion technology, we could acquire rich information from each of the precious cancer samples. The spatial data at single-cell resolution enabled us to assess the heterogeneity of tumour tissue by defining cell types, immune infiltration, and cancer-immune cell interaction within an undissociated tissue section. Future analysis and application of Hyperion data would allow us to find better predictors for colorectal cancer tissue with more accurate diagnosis and prognosis.Ethics ApprovalThis study was approved by the Institutional Review Board (#1050191) at Intermountain Healthcare (Salt Lake City, UT USA)

Uncertainty Analysis for Topographic Correction of Hyperspectral Remote Sensing Images

Quantitative uncertainty analysis is generally taken as an indispensable step in the calibration of a remote sensor. A full uncertainty propagation chain has not been established to set up the metrological traceability for surface reflectance inversed from remotely sensed images. As a step toward this goal, we proposed an uncertainty analysis method for the two typical semi-empirical topographic correction models, i.e., C and Minnaert, according to the ‘Guide to the Expression of Uncertainty in Measurement (GUM)’. We studied the data link and analyzed the uncertainty propagation chain from the digital elevation model (DEM) and at-sensor radiance data to the topographic corrected radiance. We obtained spectral uncertainty characteristics of the topographic corrected radiance as well as its uncertainty components associated with all of the input quantities by using a set of Earth Observation-1 (EO-1) Hyperion data acquired over a rugged soil surface partly covered with snow. Firstly, the relative uncertainty of cover types with lower radiance values was larger for both C and Minnaert corrections. Secondly, the trend of at-sensor radiance contributed to a spectral feature, where the uncertainty of the topographic corrected radiance was poor in bands below 1400 nm. Thirdly, the uncertainty components associated with at-sensor radiance, slope, and aspect dominated the total combined uncertainty of corrected radiance. It was meaningful to reduce the uncertainties of at-sensor radiance, slope, and aspect for reducing the uncertainty of corrected radiance and improving the data quality. We also gave some suggestions to reduce the uncertainty of slope and aspect data.

MULTISPECTRAL AND HYPERSPECTRAL SATELLITE DATA FOR ALTERATION MAPPING USING PARTIAL-SUBPIXEL ALGORITHM IN AN UNEXPLORED REGION

The focus of this paper is to comparatively evaluate the performance of ASTER and Hyperion data for target detection of hydrothermal alteration zones associated with geothermal (GT) springs in an unexplored savannah region. The study employed the partial subpixel unmixing Mixture Tuned Match Filtering algorithm for spectral information extraction using the multispectral and hyperspectral satellite data. In both cases, image endmember spectra specifically for kaolinite, alunite, and illite and calcite zones were selected and extracted by using the Analytical Imaging and Geophysics (AIG)-developed processing methods. The results show that the Hyperion, despite its distortions, can effectively discriminate associated alteration zones better than ASTER. Consequently, Hyperspectral data and analysis are thus recommended for use in similar unexplored regions for GT resource detection and monitoring.

Tree Species Classification Using Hyperion and Sentinel-2 Data with Machine Learning in South Korea and China

Remote sensing (RS) has been used to monitor inaccessible regions. It is considered a useful technique for deriving important environmental information from inaccessible regions, especially North Korea. In this study, we aim to develop a tree species classification model based on RS and machine learning techniques, which can be utilized for classification in North Korea. Two study sites were chosen, the Korea National Arboretum (KNA) in South Korea and Mt. Baekdu (MTB; a.k.a., Mt. Changbai in Chinese) in China, located in the border area between North Korea and China, and tree species classifications were examined in both regions. As a preliminary step in developing a classification algorithm that can be applied in North Korea, common coniferous species at both study sites, Korean pine (Pinus koraiensis) and Japanese larch (Larix kaempferi), were chosen as targets for investigation. Hyperion data have been used for tree species classification due to the abundant spectral information acquired from across more than 200 spectral bands (i.e., hyperspectral satellite data). However, it is impossible to acquire recent Hyperion data because the satellite ceased operation in 2017. Recently, Sentinel-2 satellite multispectral imagery has been used in tree species classification. Thus, it is necessary to compare these two kinds of satellite data to determine the possibility of reliably classifying species. Therefore, Hyperion and Sentinel-2 data were employed, along with machine learning techniques, such as random forests (RFs) and support vector machines (SVMs), to classify tree species. Three questions were answered, showing that: (1) RF and SVM are well established in the hyperspectral imagery for tree species classification, (2) Sentinel-2 data can be used to classify tree species with RF and SVM algorithms instead of Hyperion data, and (3) training data that were built in the KNA cannot be used for the tree classification of MTB. Random forests and SVMs showed overall accuracies of 0.60 and 0.51 and kappa values of 0.20 and 0.00, respectively. Moreover, combined training data from the KNA and MTB showed high classification accuracies in both regions; RF and SVM values exhibited accuracies of 0.99 and 0.97 and kappa values of 0.98 and 0.95, respectively.