Ensemble Methodology Using Multistage Learning for Improved Detection of Harmful Algal Blooms

Gokaraju, Balakrishna; Durbha, Surya S.; King, Roger L.; Younan, Nicolas H.

doi:10.1109/lgrs.2011.2182032

Cited by 16 publications

(11 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Support Vector Machines (SVMs) have been proposed for this application by Li et al [30] and Song et al [7]. Spatiotemporal analysis using machine learning methods have also been proposed by Gokaraju et al [5], [6]. Other non-machine learning methods have been proposed for HAB monitoring, detection and prediction (e.g.…”

Section: Review Of Hab Detection Methodsmentioning

confidence: 99%

“…Although difficult to compare directly, Gokaraju et al have developed limited spatiotemporal methods of HAB detection using SVMs and Neural Networks [5], [6]. These pieces of work both use ground truth in Florida but only a very small data set (less than 30 datapoints for MODIS based estimation).…”

Section: A Conventional Comparative Methodsmentioning

confidence: 99%

“…Although there is possibly redundancy between the chosen bands and products such as Chl-a, the inclusion of these bands was intended to provide fine-grain classification and discrimination for the use of this key available data. All these modalities have been used by previous spatiotemporal HAB detection methods [5], [6]. Bathymetry obviously did not vary over time, but the same modality format was used as shown where the image samples were resampled/interpolated to be co-located as the images of the other modalities.…”

Section: A Selected Modalitiesmentioning

confidence: 99%

See 2 more Smart Citations

HABNet: Machine Learning, Remote Sensing-Based Detection of Harmful Algal Blooms

Hill

Kumar

Temimi

et al. 2020

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

This paper describes the application of machine learning techniques to develop a state-of-the-art detection and prediction system for spatiotemporal events found within remote sensing data; specifically, Harmful Algal Bloom events (HABs). We propose an HAB detection system based on: a ground truth historical record of HAB events, a novel spatiotemporal datacube representation of each event (from MODIS and GEBCO bathymetry data) and a variety of machine learning architectures utilising state-of-the-art spatial and temporal analysis methods based on Convolutional Neural Networks (CNNs), Long Short-Term Memory (LSTM) components together with Random Forest and Support Vector Machine (SVM) classification methods. This work has focused specifically on the case study of the detection of Karenia Brevis Algae (K. brevis) HAB events within the coastal waters of Florida (over 2850 events from 2003 to 2018; an order of magnitude larger than any previous machine learning detection study into HAB events). The development of multimodal spatiotemporal datacube data structures and associated novel machine learning methods give a unique architecture for the automatic detection of environmental events. Specifically, when applied to the detection of HAB events it gives a maximum detection accuracy of 91% and a Kappa coefficient of 0.81 for the Florida data considered. A HAB forecast system was also developed where a temporal subset of each datacube was used to predict the presence of a HAB in the future. This system was not significantly less accurate than the detection system being able to predict with 86% accuracy up to 8 days in the future.

show abstract

Section: Review Of Hab Detection Methodsmentioning

confidence: 99%

Section: A Conventional Comparative Methodsmentioning

confidence: 99%

Section: A Selected Modalitiesmentioning

confidence: 99%

See 1 more Smart Citation

HABNet: Machine Learning, Remote Sensing-Based Detection of Harmful Algal Blooms

Hill

Kumar

Temimi

et al. 2020

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

show abstract

“…In such cases, artificial neural networks are used for binary classification of Levee change detection w.r.t backscatter reflectance [15], and [18]. Another approach includes the study of spatio-temporal contextual information for disaster events such as harmful algal blooms, where Support vector machines (SVMs) are used for temporal classification of harmful vs non-harmful algal blooms in coastal waters [16], and [19]. Many other recent data fusion studies for change detection include band ratioing and maximum likelihood classifiers, fuzzy logic and log-cumulants, Kullback-Leibler distance measure for expectation maximization, Markov random fields etc.…”

Section: B Data Fusionmentioning

confidence: 99%

Change detection analysis of tornado disaster using conditional copulas and Data Fusion for cost-effective disaster management

Gokaraju

Turlapaty

Doss

et al. 2015

2015 IEEE Applied Imagery Pattern Recognition Workshop (AIPR)

View full text Add to dashboard Cite

The up-to-date results are presented from an ongoing study of the Data Fusion of multi-temporal and multi-sensor satellite datasets for near real time damage and debris assessment after a tornado disaster event. The space-borne sensor datasets comprising of: (i) C-band SAR dataset from RADARSAT-2; (ii) Multi-Spectral (MS) optical dataset including NIR from RapidEye; (iii) MS and panchromatic dataset of Advanced Linear Imaging (ALI), are studied for multi-sensor data fusion. A combined approach of multipolarized radiometric and textural feature extraction, and statistical learning based feature classification is devised for fine tuning of the complex and generalized change detection model. We also investigated the use of multi-variate conditional copula as a classifier technique, by formulating the change and no-change as a binary-class classification problem in this study. The classification results from the above technique are used for assessment of damage and debris cover after the tornado disaster event. The performance of the above approach yields a very significant Kappa accuracy up to 75%. A 10-fold cross validation strategy is used for quantitative analysis of the performance of the classification model. This study will be further extended for modelling the effect of incidence angle discrepancies or climatic condition variances, which will address the heterogeneity factor in terms of local statistics of the dataset.

show abstract

“…Similarly, in [ 9 ], a higher order singular value decomposition (HOSVD)-based fusion approach was presented. Moreover, neural network-based evolutionary approaches for non-linear parametric identification have been proposed for data fusion [ 10 ].…”

Section: Introductionmentioning

confidence: 99%

Multi-Scale Pixel-Based Image Fusion Using Multivariate Empirical Mode Decomposition

Rehman

Ehsan

Abdullah³

et al. 2015

Sensors

View full text Add to dashboard Cite

A novel scheme to perform the fusion of multiple images using the multivariate empirical mode decomposition (MEMD) algorithm is proposed. Standard multi-scale fusion techniques make a priori assumptions regarding input data, whereas standard univariate empirical mode decomposition (EMD)-based fusion techniques suffer from inherent mode mixing and mode misalignment issues, characterized respectively by either a single intrinsic mode function (IMF) containing multiple scales or the same indexed IMFs corresponding to multiple input images carrying different frequency information. We show that MEMD overcomes these problems by being fully data adaptive and by aligning common frequency scales from multiple channels, thus enabling their comparison at a pixel level and subsequent fusion at multiple data scales. We then demonstrate the potential of the proposed scheme on a large dataset of real-world multi-exposure and multi-focus images and compare the results against those obtained from standard fusion algorithms, including the principal component analysis (PCA), discrete wavelet transform (DWT) and non-subsampled contourlet transform (NCT). A variety of image fusion quality measures are employed for the objective evaluation of the proposed method. We also report the results of a hypothesis testing approach on our large image dataset to identify statistically-significant performance differences.

show abstract

Ensemble Methodology Using Multistage Learning for Improved Detection of Harmful Algal Blooms

Cited by 16 publications

References 11 publications

HABNet: Machine Learning, Remote Sensing-Based Detection of Harmful Algal Blooms

HABNet: Machine Learning, Remote Sensing-Based Detection of Harmful Algal Blooms

Change detection analysis of tornado disaster using conditional copulas and Data Fusion for cost-effective disaster management

Multi-Scale Pixel-Based Image Fusion Using Multivariate Empirical Mode Decomposition

Contact Info

Product

Resources

About