Description: Low-cost unmanned airborne vehicles (UAVs) are emerging as a promising platform for remote-sensing data acquisition to satisfy the needs of wide range of applications. Utilizing UAVs, which are equipped with directly georeferenced RGB-frame cameras and hyperspectral push-broom scanners, for precision agriculture and high-throughput phenotyping is an important application that is gaining significant attention from researchers in the mapping and plant science fields. The advantages of UAVs as mobile-mapping platforms include low cost, ease of storage and deployment, ability to fly lower and collect high-resolution data, and filling an important gap between wheel-based and manned-airborne platforms. However, limited endurance and payload are the main disadvantages of consumer-grade UAVs. These limitations lead to the adoption of low-quality direct georeferencing and imaging systems, which in turn will impact the quality of the delivered products. Thanks to recent advances in sensor calibration and automated triangulation, accurate mapping using low-cost frame imaging systems equipped with consumer-grade georeferencing units is feasible. Unfortunately, the quality of derived geospatial information from push-broom scanners is quite sensitive to the performance of the implemented direct georeferencing unit. This paper presents an approach for improving the orthorectification of hyperspectral push-broom scanner imagery with the help of generated orthophotos from frame cameras using tie point and linear features, while modeling the impact of residual artifacts in the direct georeferencing information. The performance of the proposed approach has been verified through real datasets that have been collected by quadcopter and fixed-wing UAVs over an agricultural field.

Description: Open ocean and coastal area monitoring requires multispectral satellite images with a middle spatial resolution $({sim 300 {text{m}}})$ and a high temporal repeatability $({sim 1 {text{h}}})$ . As no current satellite sensors have such features, the aim of this study is to propose a fusion method to merge images delivered by a low earth orbit (LEO) sensor with images delivered by a geostationary earth orbit (GEO) sensor. This fusion method, called spatial spectral temporal fusion (SSTF), is applied to the future sensors—Ocean and Land Color Instrument (OLCI) (on Sentinel-3) and Flexible Combined Imager (FCI) (on Meteosat Third Generation) whose images were simulated. The OLCI bands, acquired at t 0 , are divided by the oversampled corresponding FCI band acquired at t 0 and multiplied by the FCI bands acquired at t 1 . The fusion product is used for the next fusion at t 1 and so on. The high temporal resolution of FCI allows its signal-to-noise ratio (SNR) to be enhanced by the means of temporal filtering. The fusion quality indicator ERGAS computed between SSTF fusion products and reference images is around 0.75, once the FCI images are filtered from the noise and 1.08 before filtering. We also compared the estimation of chlorophyll (Chl ) , suspended particulate matter (SPM), and colored dissolved organic matter (CDOM) maps from the fusion products with the input simulation maps. The comparison shows an average relative error s on Chl, SPM, and CDOM, respectively, of 64.6%, 6.2%, and 9.5% with the SSTF method. The SSTF method was also compared with an existing fusion method called the spati- l and temporal adaptive reflectance fusion model (STARFM).

Description: Sky/cloud images captured by ground-based cameras (a.k.a. whole sky imagers) are increasingly used nowadays because of their applications in a number of fields, including climate modeling, weather prediction, renewable energy generation, and satellite communications. Due to the wide variety of cloud types and lighting conditions in such images, accurate and robust segmentation of clouds is challenging. In this paper, we present a supervised segmentation framework for ground-based sky/cloud images based on a systematic analysis of different color spaces and components, using partial least-squares regression. Unlike other state-of-the-art methods, our proposed approach is entirely learning based and does not require any manually defined parameters. In addition, we release the Singapore whole Sky imaging segmentation database, a large database of annotated sky/cloud images, to the research community.

Description: Several methods have been proposed for atmospheric correction over turbid waters, including near-infrared (NIR) band based or short-wave infrared (SWIR) band-based where the signal in turbid waters can be assumed zero. Here, we adopt a new infrared extrapolation method to extend the existing turbid water atmospheric correction of the operational land imager (OLI) data on Landsat-8 platform. The atmospheric correction uses the extrapolated Rayleigh-corrected reflectance at NIR and SWIR bands to derive the ratios of NIR to SWIR and visible aerosol single-scattering contributions (aerosol epsilon). Taking the Pearl River Estuary as an example, the magnitude and spatial distribution of reflectance from OLI compare well with those of concurrent moderate resolution imaging spectroradiometer /Aqua based on SWIRE atmospheric correction method. The linear regression coefficients between the resampled OLI and Aqua data have demonstrated the proposed atmospheric correction method can provide robust and realistic reflectance. The advantages of the high spatial resolution made the OLI data a good source for applications in coastal and estuarine waters.

Description: Use of remotely sensed (e.g., Landsat) imagery for developing sampling frame strata for large-scale inventories of natural resources has potential for increasing sampling efficiency and lowering cost by reducing required sample sizes. Sampling frame errors are inherent with the use of this technology, either from user misclassification or due to flawed technology. Knowledge of these sampling frame errors is important, as they inflate the variance of inventory estimates, particularly poststratified estimates. Forest inventory estimates from the Mississippi Institute for Forest Inventory (MIFI) were utilized to study the extent to which Geographic Information System classification errors (sampling frame errors) affect forest volume and area mean and variance estimates. MIFI's high sampling intensity provided a unique opportunity to quantify the magnitude that different levels of misclassification ultimately have on mean and variance estimates. A variance calculator was developed to assess the impact of various levels of misclassification on least and most variable summary estimates of cubic meter volume percent and total area. The standard error estimates for mean and total volume decreased when plots were reallocated to their correct strata. The increased efficiency obtained from correcting misclassifications illustrates that the loss in precision due to misclassifying inventory strata is consequential. Knowledge and correction of these errors provides a natural-resource-based professional or investor using land classification/inventory data the best minimum risk information possible. A complete set of variance estimators for poststratified means and total area estimates with sampling frame errors are presented and compared to estimators without sampling frame errors.

Description: In this paper, a graph regularized nonlinear ridge regression (RR) model is proposed for remote sensing data analysis, including hyper-spectral image classification and atmospheric aerosol retrieval. The RR is an efficient linear regression method, especially in handling cases with a small number of training samples or with correlated features. However, large amounts of unlabeled samples exist in remote sensing data analysis. To sufficiently explore the information in unlabeled samples, we propose a graph regularized RR (GRR) method, where the vertices denote labeled or unlabeled samples and the edges represent the similarities among different samples. A natural assumption is that the predict values of neighboring samples are close to each other. To further address the nonlinearly separable problem in remote sensing data caused by the complex acquisition process as well as the impacts of atmospheric and geometric distortions, we extend the proposed GRR into a kernelized nonlinear regression method, namely KGRR. To evaluate the proposed method, we apply it to both classification and regression tasks and compare with representative methods. The experimental results show that KGRR can achieve favorable performance in terms of predictability and computation time.

Description: Area-to-point regression kriging (ATPRK) is an advanced image fusion approach in remote sensing In this paper, ATPRK is considered for sharpening hyperspectral images (HSIs), based on the availability of a fine spatial resolution panchromatic or multispectral image. ATPRK can be used straightforwardly to sharpen each coarse hyperspectral band in turn. This scheme, however, is computationally expensive due to the large number of bands in HSIs, and this problem is exacerbated for multiscene or multitemporal analysis. Thus, we extend ATPRK for fast HSI sharpening with a new approach, called approximate ATPRK (AATPRK), which transforms the original HSI to a new feature space and image fusion is performed for only the first few components before back transformation. Experiments on two HSIs show that AATPRK greatly expedites ATPRK, but inherits the advantages of ATPRK, including maintaining a very similar performance in sharpening (both ATPRK and AATPRK can produce more accurate results than seven benchmark methods) and precisely conserving the spectral properties of coarse HSIs.

Description: A firefly algorithm (FA) inspired band selection and optimized extreme learning machine (ELM) for hyperspectral image classification is proposed. In this framework, FA is to select a subset of original bands to reduce the complexity of the ELM network. It is also adapted to optimize the parameters in ELM (i.e., regularization coefficient C , Gaussian kernel σ, and hidden number of neurons L ). Due to very low complexity of ELM, its classification accuracy can be used as the objective function of FA during band selection and parameter optimization. In the experiments, two hyperspectral image datasets acquired by HYDICE and HYMAP are used, and the experiment results indicate that the proposed method can offer better performance, compared with particle swarm optimization and other related band selection algorithms.

Description: Simplex growing algorithm (SGA) is an endmember finding algorithm which searches for endmembers one after another by growing simplexes one vertex at a time via maximizing simplex volume (SV). Unfortunately, several issues arise in calculating SV. One is the use of dimensionality reduction (DR) because the dimensionality of a simplex is usually much smaller than data dimensionality. Second, calculating SV requires calculating the determinant of an ill-ranked matrix in which case singular value decomposition (SVD) is generally required to perform DR. Both approaches generally do not produce true SV. Finally, the computing time becomes excessive and numerically instable as the number of endmembers grows. This paper develops a new theory, called geometric simplex growing analysis (GSGA), to resolve these issues. It can be considered as an alternative to SGA from a rather different point of view. More specifically, GSGA looks into the geometric structures of a simplex whose volume can be actually calculated by multiplication of its base and height. As a result, it converts calculating maximal SV to finding maximal orthogonal projection as its maximal height becomes perpendicular to its base. To facilitate GSGA in practical applications, GSGA is further used to extend SGA to recursive geometric simplex growing algorithm (RGSGA) which allows GSGA to be implemented recursively in a similar manner that a Kalman filter does. Consequently, RGSGA can be very easily implemented with significant saving of computing time. Best of all, RGSGA is also shown to be most efficient and effective among all SGA-based variants.

Description: The localization of changes that occur between the images in a multitemporal series is crucial for many applications, ranging from environmental monitoring to military surveillance. In contrast to traditional change detection methods, unmixing-based change detection has been shown to have the important added benefit of providing subpixel-level information on the nature of the changes, instead of only providing the location of the changes. Recently, sparse unmixing has also been introduced to hyperspectral change detection, resulting in a method that circumvents the drawbacks of regular spectral unmixing approaches. Sparse unmixing-based change detection reveals the changes that occur in a multitemporal series, at subpixel level, and in terms of the library spectra and their sparse abundances, and provides enhanced change detection performance, especially when subpixel-level changes have occurred. However, sparse unmixing is generally an ill-conditioned and time-consuming process, especially as the size of the utilized spectral library increases. In this paper, dictionary pruning is exploited for the first time for hyperspectral change detection using sparse unmixing, in order to alleviate the ill-conditioning of the problem and achieve decreased computation times and enhanced change detection performance. Experimental results on both realistic synthetic and real datasets are used to validate the proposed approach.