Filters








2,104 Hits in 5.2 sec

High-performance computing in remotely sensed hyperspectral imaging: the Pixel Purity Index algorithm as a case study

A. Plaza, D. Valencia, J. Plaza
2006 Proceedings 20th IEEE International Parallel & Distributed Processing Symposium  
This paper explores three HPC-based paradigms for efficient information extraction from remote sensing data using the Pixel Purity Index (PPI) algorithm (available from the popular Kodak's Research Systems  ...  For instance, hyperspectral imaging is a new technique in remote sensing that generates hundreds of spectral bands at different wavelength channels for the same area on the surface of the Earth.  ...  As a case study, this paper focuses on the pixel purity index (PPI) algorithm, one of the most widely used standard algorithms in the hyperspectral imaging community.  ... 
doi:10.1109/ipdps.2006.1639607 dblp:conf/ipps/PlazaVP06 fatcat:zy4gbzznlvhbdohyjs5ok2uslu

FPGA implementation of endmember extraction algorithms from hyperspectral imagery: pixel purity index versus N-FINDR

Carlos Gonzalez, Daniel Mozos, Javier Resano, Antonio Plaza, Bormin Huang, Antonio J. Plaza
2011 High-Performance Computing in Remote Sensing  
In this paper, we perform an inter-comparison of the hardware implementations of two widely used techniques for automatic endmember extraction from remotely sensed hyperspectral images: the pixel purity  ...  index (PPI) and the N-FINDR.  ...  ACKNOWLEDGMENTS This work has been supported by the European Community's Marie Curie Research Training Networks Programme under reference MRTN-CT-2006-035927, Hyperspectral Imaging Network (HYPER-I-NET  ... 
doi:10.1117/12.897384 fatcat:efqqxtpbsrhdlp7sgdr7rz57pe

QuasarNET: Human-level spectral classification and redshifting with Deep Neural Networks [article]

Nicolas Busca, Christophe Balland
2018 arXiv   pre-print
We introduce QuasarNET, a deep convolutional neural network that performs classification and redshift estimation of astrophysical spectra with human-expert accuracy.  ...  We pose these two tasks as a feature detection problem: presence or absence of spectral features determines the class, and their wavelength determines the redshift, very much like human-experts proceed  ...  learning techniques.  ... 
arXiv:1808.09955v1 fatcat:rio7ljxjf5hk5oep2gygthr7cy

Parallel Implementation of Hyperspectral Image Processing Algorithms

A. Plaza, D. Valencia, J. Plaza, J. Sanchez-Testal, S. Munoz, S. Blazquez
2006 2006 IEEE International Symposium on Geoscience and Remote Sensing  
In this paper, we take a necessary first step towards the quantitative comparison of parallel techniques and strategies for analyzing hyperspectral data sets.  ...  Our focus is on three types of algorithms: automatic target recognition, spectral mixture analysis and data compression.  ...  Every pixel in the input data is projected onto each skewer, and the number of times a given pixel is selected as extreme defines its purity index.  ... 
doi:10.1109/igarss.2006.242 dblp:conf/igarss/PlazaVPSMB06 fatcat:3hyv7i362jdf3fdc4dqa4slrxe

Incorporation of spatial constraints into spectral mixture analysis of remotely sensed hyperspectral data

Antonio Plaza, Javier Plaza, Gabriel Martin
2009 2009 IEEE International Workshop on Machine Learning for Signal Processing  
This approach involves the separation of a mixed pixel into its pure components or endmember spectra, and the estimation of the abundance value for each endmember.  ...  For this purpose, we discuss the advantages that can be obtained after including spatial information in techniques for endmember extraction and fractional abundance estimation, using a database of synthetic  ...  Several approaches have been developed for automatic or semi-automatic endmember extraction, including the pixel purity index (PPI) algorithm [5] , the orthogonal subspace projection (OSP) [6] , the  ... 
doi:10.1109/mlsp.2009.5306202 fatcat:sb5mei44cjgiraggrdquuk767u

Experimental Comparative Study on Autoencoder Performance for Aided Melanoma Skin Disease Recognition

Zahraa Diame, Maryam ElBery, Mohammed Salem, Mohamed Roushdy
2022 International Journal of Intelligent Computing and Information Sciences  
So, automatic skin segmentation is considered an enthusiastic study at present.  ...  Melanoma is a dangerous and metastatic cancer that may be fatal and it has a high ability to invade other tissues and organs.  ...  The training images in the data set are often of different sizes and pixels, but the deep neural network model requires input images of a fixed size.  ... 
doi:10.21608/ijicis.2022.104799.1136 fatcat:icz5utcqrvg2nn5fs7xcidsmay

Convolutional neural networks for transient candidate vetting in large-scale surveys

Fabian Gieseke, Steven Bloemen, Cas van den Bogaard, Tom Heskes, Jonas Kindler, Richard A. Scalzo, Valério A. R. M. Ribeiro, Jan van Roestel, Paul J. Groot, Fang Yuan, Anais Möller, Brad E. Tucker
2017 Monthly notices of the Royal Astronomical Society  
We show that relatively simple networks are already competitive with state-of-the-art systems and that their quality can further be improved via slightly deeper networks and additional preprocessing steps  ...  In particular, our best model correctly classifies about 97.3% of all 'real' and 99.7% of all 'bogus' instances on a test set containing 1,942 'bogus' and 227 'real' instances in total.  ...  For this reason, automatic detection algorithms that yield both a high purity and a high completeness will play an essential role for future transient surveys.  ... 
doi:10.1093/mnras/stx2161 fatcat:m7irhzykgvfq7e67i6ubntk35e

Parallel Implementation of Endmember Extraction Algorithms From Hyperspectral Data

A. Plaza, D. Valencia, J. Plaza, C.-I. Chang
2006 IEEE Geoscience and Remote Sensing Letters  
In particular, we develop parallel implementations of approximate versions of the N-FINDR and pixel purity index algorithms, along with a parallel hybrid of both techniques.  ...  However, the intrinsic properties of available techniques are amenable to the design of parallel implementations.  ...  In particular, we develop parallel implementations of approximate versions of the N-FINDR and pixel purity index algorithms, along with a parallel hybrid of both techniques.  ... 
doi:10.1109/lgrs.2006.871749 fatcat:o6xdwg5a3ncqjflelsvbhnc6wm

Automatic selection of informative samples for SVM-based classification of hyperspectral data using limited training sets

Antonio Plaza, Javier Plaza
2010 2010 2nd Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing  
Index Terms-Machine learning, hyperspectral image classification, support vector machines (SVMs), automatic selection of training samples 978-1-4244-8907-7/10/$26.00 ©2010 IEEE  ...  This issue is investigated by comparing different unsupervised algorithms which account for the spectral purity of training samples in the process of selecting those samples for classification purposes  ...  Selection of pure training samples In order to extract the purest pixel vectors in the data set as training samples, we use a modified version of the pixel purity index (PPI) algorithm available commercially  ... 
doi:10.1109/whispers.2010.5594931 dblp:conf/whispers/PlazaP10 fatcat:mm2j6rnrnfes3pj6j4urokxeda

A texture-based pixel labeling approach for historical books

Maroua Mehri, Petra Gomez-Krämer, Pierre Héroux, Alain Boucher, Rémy Mullot
2015 Pattern Analysis and Applications  
Thus, in this article a framework is proposed to investigate the use of texture as a tool for automatically determining homogeneous regions in a digitized historical book and segmenting its contents by  ...  Over the last few years, there has been tremendous growth in the automatic processing of digitized historical documents.  ...  A widely used class of probabilistic models is: Conditional Random Fields (CRF) [43] , Markov Random Fields (MRF) [44] , Gaussian Markov Random Fields (GMRF) [45] , fractals [46] and Local Binary  ... 
doi:10.1007/s10044-015-0451-9 fatcat:b5njvrjrzjf75iq4t6lvzbqu2m

Spatial-Spectral Preprocessing Prior to Endmember Identification and Unmixing of Remotely Sensed Hyperspectral Data

Gabriel Martin, Antonio Plaza
2012 IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing  
The method first derives a spatial homogeneity index for each pixel in the hyperspectral image. This index is relatively insensitive to the noise present in the data.  ...  A challenging problem is how to automatically identify endmembers, as the presence of mixed pixels generally prevents the localization of pure spectral signatures in transition areas between different  ...  The spectral purity index for a given pixel will be the sum of all the weights for that pixel over the first principal components.  ... 
doi:10.1109/jstars.2012.2192472 fatcat:j2ijgtwx3rhddej3fu7bvqr7xq

Hyperspectral Data Processing Algorithms [chapter]

Antonio Plaza, Javier Plaza, Gabriel Martín, Sergio Sánchez
2018 Fundamentals, Sensor Systems, Spectral Libraries, and Data Mining for Vegetation  
In most studies, techniques are divided into full-pixel and mixed-pixel techniques, where each pixel vector defines a spectral signature or fingerprint that uniquely characterizes the underlying materials  ...  Mostly based on previous efforts in multispectral imaging, full-pixel techniques assume that each pixel vector measures the response of one single underlying material.  ...  Last but not least, the authors would like to take this opportunity to gratefully acknowledge the editors of this volume for their very kind invitation to contribute a chapter and for all their support  ... 
doi:10.1201/9781315164151-11 fatcat:gm3twsueujh3jla6wzuo2piamq

Evaluation of the Risk of Recurrence in Patients with Local Advanced Rectal Tumours by Different Radiomic Analysis Approaches

Alaa Khadidos, Adil Khadidos, Olfat M. Mirza, Tawfiq Hasanin, Wegayehu Enbeyle, Abdulsattar Abdullah Hamad, Fahd Abd Algalil
2021 Applied Bionics and Biomechanics  
Classically, researchers in this field of radiomics have used conventional machine learning techniques (random forest, for example).  ...  In deep learning, we built a 16-layer convolutional neural network model, driven by a 2D MRI image database comprising both the native images and the bounding box corresponding to each image.  ...  In this context, we opted for a combinatorial technique, Figure 2 , using both a selection algorithm (recursive feature elimination or RFE) with a classification algorithm (random forest or RF).  ... 
doi:10.1155/2021/4520450 pmid:34876924 pmcid:PMC8645400 fatcat:xsho5t7oarholcbyhq4fyzebmy

A Self-Supervised Learning Framework for Road Centerline Extraction From High-Resolution Remote Sensing Images

Qing Guo, Zhipan Wang
2020 IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing  
Then, the one-class classifier framework is introduced and the random forest positive unlabeled learning classifier is constructed to get the posterior probability of the pixel belonging to road.  ...  However, the deep learning method needs a large number of samples and a long training time, and our self-supervised learning framework does not need the training samples.  ...  The optimal values of all three indexes are equal to 1. The accuracy of road extraction result is not just dependent on one index, but relies on all three indexes at the same time. A.  ... 
doi:10.1109/jstars.2020.3014242 fatcat:u6ejjszqgnawndrv2d2zdmddgq

Thermal Change Index-Based Diabetic Foot Thermogram Image Classification Using Machine Learning Techniques

Amith Khandakar, Muhammad Chowdhury, Mamun Reaz, Sawal Ali, Tariq Abbas, Tanvir Alam, Mohamed Ayari, Zaid Mahbub, Rumana Habib, Tawsifur Rahman, Anas Tahir, Ahmad Bakar (+1 others)
2022 Sensors  
In this work, a publicly available dataset was categorized into different classes, which were corroborated by domain experts, based on a temperature distribution parameter—the thermal change index (TCI  ...  Classical machine learning algorithms with feature engineering and the convolutional neural network (CNN) with image enhancement techniques were extensively investigated to identify the best performing  ...  Random Forests are often used for feature selection in machine learning because the tree-based strategies used by random forests naturally rank by how well they improve the purity of the node.  ... 
doi:10.3390/s22051793 pmid:35270938 pmcid:PMC8915003 fatcat:6zq5yotfb5abjped5f7zi5i3n4
« Previous Showing results 1 — 15 out of 2,104 results