Optimizing Machine Learning Algorithms for Hyperspectral Very Shallow Water (VSW) Products

HyperSpectral Imagery (HSI) of the coastal zone often focuses on the estimation of bathymetry. However, the estimation of bathymetry requires knowledge, or the simultaneous solution, of water column Inherent Optical Properties (IOPs) and bottom reflectance. The numerical solution to the simultaneous set of equations for bathymetry, IOPs, and bottom reflectance places high demands on the spectral quality, calibration, atmospheric correction, and Signal-to-Noise (SNR) of the HSI data stream. In October of 2002, a joint FERI/NRL/NAVO/USACE HSI/LIDAR experiment was conducted off of Looe Key, FL. This experiment yielded high quality HSI data at a 2 m resolution and bathymetric LIDAR data at a 4 m resolution. The joint data set allowed for the advancement and validation of a previously generated Look-Up-Table (LUT) approach to the simultaneous retrieval of bathymetry, IOPs, and bottom type. Bathymetric differences between the two techniques were normally distributed around a 0 mean, with the exception of two peaks. One peak related to a mechanical problem in the LIDAR detector mirrors that causes errors on the edges of the LIDAR flight lines. The other significant difference occurred in a single geographic area (Hawk Channel) suggesting an incomplete IOP or bottom reflectance description in the LUT data base. In addition, benthic habitat data from NOAA’s National Ocean Service (NOS) and the Florida Wildlife Research Institute (FWRI) provided validation data for the estimation of bottom type. Preliminary analyses of the bottom type estimation suggest that the best retrievals are for seagrass bottoms. One source of the potential difficulties may be that the LUT database was generated from a more pristine location (Lee Stocking Island, Bahamas). It is expected that fusing the HSI/LIDAR data streams should reduce the errors in bottom typing and IOP estimation.

This article proposes a symmetric sparse representation (SSR) method to extract pure endmembers from hyperspectral imagery (HSI). The SSR combines the features of the linear unmixing model and the sparse subspace clustering model of endmembers, and it assumes that the desired endmembers and all the HSI pixel points can be sparsely represented by each other. It formulates the endmember extraction problem into a famous program of archetypal analysis, and accordingly, extracting pure endmembers can be transformed as finding the archetypes in the minimal convex hull containing all the HSI pixel points. The vector quantization scheme is adopted to help in carefully choosing the initial pure endmembers, and the archetypal analysis program is solved using the simple projected gradient algorithm. Seven state-of-the-art methods are implemented to make comparisons with the SSR on both synthetic and real hyperspectral images. Experimental results show that the SSR outperforms all the seven methods in spectral angle distance and root-mean-square error, and it can be a good alternative choice for extracting pure endmembers from HSI data.

A number of institutions, including the Naval Research Laboratory (NRL), have developed look up tables for remote retrieval of bathymetry and in-water optical properties from hyperspectral imagery (HSI) [6]. For bathymetry retrieval, the lower limit is the very shallow water case (here defined as < 2m), a depth zone which is not well resolved by many existing bathymetric LIDAR sensors, such as SHOALS [4]. The ability to rapidly model these shallow water depths from HSI directly has potential benefits for combined HSI/LIDAR systems such as the Compact Hydrographic Airborne Rapid Total Survey (CHARTS) [10]. In this study, we focused on the validation of a near infra-red feature, corresponding to a local minimum in absorption (and therefore a local peak in reflectance), which can be correlated directly to bathymetry with a high degree of confidence. Compared to other VNIR wavelengths, this particular near-IR feature corresponds to a peak in the correlation with depth in this very shallow water regime, and this is a spectral range where reflectance depends primarily on water depth (water absorption) and bottom type, with suspended constituents playing a secondary role.

The amount of hyperspectral imagery (HSI) data currently available is relatively small compared to other imaging modalities, and what is suitable for developing, testing, and evaluating spatial-spectral algorithms is virtually nonexistent. In this work, a significant amount of coincident airborne hyperspectral and high spatial resolution panchromatic imagery that supports the advancement of spatial-spectral feature extraction algorithms was collected to address this need. The imagery was collected in April 2013 for Ohio University by the Civil Air Patrol, with their Airborne Real-time Cueing Hyperspectral Enhanced Reconnaissance (ARCHER) sensor. The target materials, shapes, and movements throughout the collection area were chosen such that evaluation of change detection algorithms, atmospheric compensation techniques, image fusion methods, and material detection and identification algorithms is possible. This paper describes the collection plan, data acquisition, and initial analysis of the collected imagery.

Inspired by the recent rapid progress of l1-norm minimization techniques and the great success of sparse dictionary learning in image modeling, this paper proposes a sparse multi-way models clustering fusion technique to improve the classification performance in hyperspectral imagery. Multi-way models consider hyperspectral imagery data as a whole entity to treat jointly spatial and spectral modes. The whole clustering fusion method is composed three steps. Firstly, the complete hyperspectral data is grouped into several independent sub-band data sources. Then, sparse multi-way model is used to feature extraction in every band set, and divide the scene into a series of homomorphic regions. At last, we propose a fusion method to combine the information provided by each band set, it can acquire approximate supervised classification performance (such as K-nearest Neighbor classifier).The experimental results on the HYDICE imagery demonstrate the efficiency and superiority of the proposed clustering method to the classical K-means clustering method.

Hyperspectral Imagery (HSI) data are inherently more expensive and difficult to acquire than Multispectral Imagery (MSI) data. Most HSI acquisitions are currently limited to relatively low spatial resolution with the inherent limitation that they are not able to spatially resolve many important targets. Higher spatial resolution MSI data are available, however, they typically don’t provide sufficient spectral resolution to allow unambiguous identification of specific targets. We describe a new method using predictive modeling of combined HSI/MSI data to generate simulated high-spatial- resolution, high-spectral-resolution image data. Targets observed by the low spatial resolution HSI sensor are spectrally resolved through linear spectral unmixing, but not spatially resolved. Targets observed by higher spatial resolution MSI sensor are partially spatially resolved but lower spectral resolution doesn’t allow target identification. The fused data are used to predict the subpixel locations of HSI-identified subpixel targets within the superior MSI spatial resolution. Comparison of individual HSI endmember spectra to MSI signatures at multiple MSI pixels allows mapping the spatial locations of specific HSI endmembers creating a simulated HSI dataset that allows location of spectrally resolved targets. The combined data incorporate the best features of HSI/MSI to allow improved target detection and characterization.

The fusion of multiple Computer Aided Detection/Computer Aided Classification (CAD/CAC) algorithms has been shown to be effective in reducing the false alarm rate associated with the automated classification of bottom mine-like objects when applied to side-scan sonar images taken in Very Shallow Water (VSW) environments [C.M. Ciany et al., 2001],[C.M. Ciany et al., 2000]. The fusion of CAD/CAC algorithms from Raytheon and NSWC Coastal Systems Station (CSS) also has been demonstrated in the shallow water environment on a single sonar data set [C.M. Ciany et al., 2002]. This paper extends the shallow water CAD/CAC/Fusion performance analysis to an additional set of sonar data taken in the Gulf of Mexico during June 1998, and adds the outputs of a third CAD/CAC algorithm from Lockheed Martin to the fusion processing. The fusion algorithm accepts the classification confidence levels and associated contact locations from the three different CAD/CAC algorithms, clusters the contacts based on the distance between their locations, and then declares a valid target when a clustered contact passes a prescribed fusion criterion. Four different fusion criteria are evaluated: the first based on the Fisher Discriminant, the second and third based on simple and constrained binary combinations of the multiple CAD/CAC processor outputs, and the fourth based on a constrained optimization approach that minimizes the total number of false alarms over the clustering distance and cluster confidence factor thresholds for a given probability of correct classification. The resulting performance of the four fusion algorithms is compared, and the overall performance benefit of a significant reduction of false alarms at high correct classification probabilities is quantified.

Research has shown that naturally occurring light outdoors and underwater is partially linearly polarized. The polarized components can be combined to form an image that describes the polarization of the light in the scene. This image is known as the degree of linear polarization (DOLP) image or partial polarization image. These naturally occurring polarization signatures can provide a diver or an unmanned underwater vehicle (UUV) with more information to detect, classify, and identify threats such as obstacles and/or mines in the shallow water environment. The SHallow water Real-time IMaging Polarimeter (SHRIMP), recently developed under sponsorship of Dr. Tom Swean at the Office of Naval Research (Code 321OE), can measure underwater partial polarization imagery. This sensor is a passive, three-channel device that simultaneously measures the three components of the Stokes vector needed to determine the partial linear polarization of the scene. The testing of this sensor has been completed and the data has been analyzed. This paper presents performance results from the field-testing and quantifies the gain provided by the partial polarization signature of targets in the Very Shallow Water (VSW) and Surf Zone (SZ) regions.

The fusion of hyperspectral imagery (HSI) and light detection and ranging (LiDAR) data is widely used for land cover classification. However, due to different imaging mechanisms, HSI and LiDAR data always present significant image differences, and the dimensions and feature distributions of HSI and LiDAR are highly dissimilar. This makes it challenging to represent and correlate semantic information from multimodal data. Current methods for classifying pixel-by-pixel features, which rely on cascaded or attention-based fusion, cannot effectively use multimodal features. To achieve accurate classification results, extracting and fusing similar high-order semantic information and complementary discriminative information contained in multimodal data is vital. In this paper, we propose a cross-modal semantic enhancement network (CMSE) for multimodal semantic information mining and fusion. Our proposed CMSE framework extracts features from the image on multiple scales, capturing more representative local sparse features with different sizes of convolution kernels. To represent high-level semantic features related to land cover, we establish a Gaussian-weighted matrix and semantically transform the spatial and spectral features of distinct branches. Finally, we build a multi-level residual fusion module to incrementally fuse spectral features from HSI and elevation features from LiDAR. Additionally, we introduce a cross-modal semantically constrained loss to guide multimodal semantic feature alignment. We evaluate our approach on three multimodal remote sensing datasets, namely the Houston2013, Trento, and MUUFL datasets. The experimental results demonstrate that our proposed CMSE model achieves superior performance in terms of accuracy and robustness compared to other related deep networks.

Due to the data-hungry nature of deep neural networks and the high dimensionality of hyperspectral imagery (HSI) data, the scarcity of training samples remains a significant challenge for HSI classification based on deep learning. Recently, the pairing or recombining of samples has proven to be an efficient approach to increase the amount of input training data, and accordingly, the architecture of a network trained by these samples has to be redesigned, since the input is no longer a single sample. Following this strategy, in this article, we propose a spatial–spectral relation network (SS-RN) for HSI classification with limited training samples. The SS-RN takes advantage of the relation network architecture to precisely capture the profound similarity between samples. Additionally, to make SS-RN more suitable for HSI classification, an entire three-dimensional (3-D) neighborhood instead of the isolated pixel is considered to explore both spectral and spatial information thoroughly. Multiple support samples are also selected for each class to make sure that the extracted features are stable enough to avoid the intraclass similarity and interclass dissimilarity problem of HSI. The network is composed on the basis of 3-D convolutional neural network blocks with the aim of extracting spatial–spectral features. Extensive experiments on three widely used HSI datasets demonstrate that the proposed method can achieve better classification accuracy than conventional deep learning methods with limited labeled training samples.

Bioluminescence Potential in the Transition Zone to Very Shallow Water (VSW)

Traditional soil nutrient and moisture analysis is labor-intensive, costly, and lacks spatial granularity, hindering precision agriculture. This paper proposes an integrated machine learning pipeline that fuses hyperspectral imagery (HSI) a powerful environmental sensing technology with advanced computer vision techniques to address this gap. We hypothesize that deep learning models can decode the complex, non-linear spectral signatures in HSI data to predict key soil properties accurately. Our methodology encompasses HSI data preprocessing, feature extraction using a Convolutional Neural Network (CNN), and regression modeling. Using a public dataset, we demonstrate that our proposed CNN-based model outperforms traditional spectral indices and machine learning models like Support Vector Regression (SVR) in predicting soil organic carbon (SOC) and moisture content. The results indicate the high potential of this pipeline for generating high-resolution, actionable soil maps to optimize resource use in agriculture.

: Aircraft and satellite Remote Sensing (RS) platforms provide spatial and temporal coverage of oceanic water conditions that are unobtainable by any other cost effective means. The hope of HyperSpectral Imagery (HSI) data is that it will provide the necessary data stream to simultaneously describe the atmospheric and water column optical properties. The goal of these hyperspectral programs is to develop the instruments, platforms, and data analysis techniques to achieve the depth-dependent description of atmospheric and water column Inherent Optical Properties (IOPs).

In order to overcome low detection accuracy and high false alarm rate of anomaly target detection based on Kernel RX anomaly detector in hyperspectral imagery (HSI). The paper proposes a new HSI anomaly target detection algorithm based on optimal bands subspace that analyses the characteristics of HSI data. First, the wavelet transform method is used to denoise HSI, which makes that the background information of hyperspectral imagery is suppressed by using the advantages of wavelet transform. Then, HSI is divided for band subsets by the classical adaptive subspace method. The optimal bands are selected by using the minimum variance index in each band subsets. Finally, the Kernel RX algorithm is used to detect anomaly target from optimal band subspaces. The data sets of real AVIRIS and synthetic HSI are used in the experiments, and the results show that the proposed algorithm has strong robustness and lower false alarm probability.

The hyperspectral/spatial detection of edges (HySPADE) algorithm, originally published in 2004 [1], has been modified and applied to a wider diversity of hyperspectral imagery (HSI) data. As originally described in [1], HySPADE operates by converting the naturally two-dimensional edge detection process based on traditional image analysis methods into a series of one-dimensional edge detections based on spectral angle. The HySPADE algorithm: i) utilizes spectral signature information to identify edges; ii) requires only the spectral information of the HSI scene data and does not require a spectral library nor spectral matching against a library; iii) facilitates simultaneous use of all spectral information; iv) does not require endmember or training data selection; v) generates multiple, independent data points for statistical analysis of detected edges; vi) is robust in the presence of noise; and vii) may be applied to radiance, reflectance, and emissivity data--though it is applied to radiance and reflectance spectra (and their principal components transformation) in this report. HySPADE has recently been modified to use Euclidean distance values as an alternative to spectral angle. It has also been modified to use an N x N-pixel sliding window in contrast to the 2004 version which operated only on spatial subset image chips. HySPADE results are compared to those obtained using traditional (Roberts and Sobel) edge-detection methods. Spectral angle and Euclidean distance HySPADE results are superior to those obtained using the traditional edge detection methods; the best results are obtained by applying HySPADE to the first few, information-containing bands of principal components transformed data (both radiance and reflectance). However, in practice, both the Euclidean distance and spectral angle versions of HySPADE should be applied and their results compared. HySPADE results are shown; extensions of the HySPADE concept are discussed as are applications for HySPADE in HSI analysis and exploitation.