Pedro Carmona - Academia.edu (original) (raw)

Papers by Pedro Carmona

Research paper thumbnail of ICPRAM 2012 - Proceedings of the 1st International Conference on Pattern Recognition Applications and Methods, Volume 1, Vilamoura, Algarve, Portugal, 6-8 February, 2012

Research paper thumbnail of Non-invasive Melanoma Diagnosis using Multispectral Imaging

The early analysis of pigmented skin lesions is important for clinicians in order to recognize ma... more The early analysis of pigmented skin lesions is important for clinicians in order to recognize malignant melanoma. However, it is difficult to differentiate it from benign skin lesions due to their similarity based on their appearance. Since melanoma has a tendency to grow inside the skin and the depth of penetration of light into the skin is wavelength dependent, a multispectral imaging acquisition and processing approach to classify pigmented lesions as melanoma seems appropriate. This paper presents a method to diagnose melanoma lesions over a group of 26 samples acquired with a multispectral system, where 6 of them are melanomas, and the other 20 are other types of pigmented lesions. A Leave-One-Out strategy is used to create the training/test set. The classification imbalance problem inherent to this dataset is alleviated using a SMOT E technique. The random component of the SMOT E methodology is dealt with running it 25 times and a Qualified Majority Voting (QMV ) scheme is used to do the final classification, using SV M. Results show this strategy allows to obtain competitive classification quality results.

Research paper thumbnail of Security authentication using phase-encoded nanoparticle structures and polarized light

Optics Letters, 2015

Phase-encoded nanostructures such as quick response (QR) codes made of metallic nanoparticles are... more Phase-encoded nanostructures such as quick response (QR) codes made of metallic nanoparticles are suggested to be used in security and authentication applications. We present a polarimetric optical method able to authenticate random phase-encoded QR codes. The system is illuminated using polarized light, and the QR code is encoded using a phase-only random mask. Using classification algorithms, it is possible to validate the QR code from the examination of the polarimetric signature of the speckle pattern. We used Kolmogorov-Smirnov statistical test and Support Vector Machine algorithms to authenticate the phase-encoded QR codes using polarimetric signatures.

Research paper thumbnail of Feature selection in regression tasks using conditional mutual information

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2011

This paper presents a supervised feature selection method applied to regression problems. The sel... more This paper presents a supervised feature selection method applied to regression problems. The selection method uses a Dissimilarity matrix originally developed for classification problems, whose applicability is extended here to regression and built using the conditional mutual information between features with respect to a continuous relevant variable that represents the regression function.

Research paper thumbnail of Effect of Denoising in Band Selection for Regression Tasks in Hyperspectral Datasets

IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2013

ABSTRACT This paper presents a comparative analysis of six band selection methods applied to hype... more ABSTRACT This paper presents a comparative analysis of six band selection methods applied to hyperspectral datasets for biophysical variable estimation problems, where the effect of denoising on band selection performance has also been analyzed. In particular, we consider four hyperspectral datasets and three regressors of different nature (ε-SVR, Regression Trees, and Kernel Ridge Regression). Results show that the denoising approach improves the band selection quality of all the tested methods. We show that noise filtering is more beneficial for the selection methods that use an estimator based on the whole dataset for the prediction of the output than for methods that use strategies based on local information (neighboring points).

Research paper thumbnail of Band selection in spectral imaging for classification and regression tasks using information theoretic measures

2011 10th Euro-American Workshop on Information Optics, 2011

In this paper we present three different methodologies of band selection for hyperspectral data s... more In this paper we present three different methodologies of band selection for hyperspectral data sets applied to classification and regression tasks using Information Theory measures. In one of the cases, the bands will be selected having information about the classification labels of the data points (supervised classification). In the second one, no information about the target labels is required (unsupervised classification). In the third problem, the target variables are of continuous nature and are also available (supervised regression).

Research paper thumbnail of Band selection in spectral imaging for non-invasive melanoma diagnosis

Biomedical Optics Express, 2013

A method consisting of the combination of the Synthetic Minority Over-Sampling TEchnique (SMOTE) ... more A method consisting of the combination of the Synthetic Minority Over-Sampling TEchnique (SMOTE) and the Sequential Forward Floating Selection (SFFS) technique is used to do band selection in a highly imbalanced, small size, two-class multispectral dataset of melanoma and non-melanoma lesions. The aim is to improve classification rate and help to identify those spectral bands that have a more important role in melanoma detection. All the processing steps were designed taking into account the low number of samples in the dataset, situation that is quite common in medical cases. The training/test sets are built using a Leave-One-Out strategy. SMOTE is applied in order to deal with the imbalance problem, together with the Qualified Majority Voting scheme (QMV). Support Vector Machines (SVM) is the classification method applied over each balanced set. Results indicate that all melanoma lesions are correctly classified, using a low number of bands, reaching 100% sensitivity and 72% specificity when considering nine (out of a total of 55) spectral bands.

Research paper thumbnail of One-class classification techniques in image recognition problems

2013 12th Workshop on Information Optics (WIO), 2013

ABSTRACT The one class classification problem is different from the binary/multi-class classifica... more ABSTRACT The one class classification problem is different from the binary/multi-class classification problem in the sense that the negative class is not present or not properly character-ized. The problem of classifying positive (target) cases in the absence of appropriately defined negative cases (outliers) has been given an increased attention during the last few years. In this paper we give a brief overview about the methods used for one class classification applied in three fields of interest: (a) remote sensing, (b) 3D image retrieval and modeling and (c) medical image/data analysis/processing. We also present results of the application of one of the methods (one-class SYM) for the detection of vegetated areas in a hyperspectral remote sensing image.

Research paper thumbnail of PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES

Research paper thumbnail of Canopy spectral invariants for remote sensing and model applications

Remote Sensing of Environment, 2007

The concept of canopy spectral invariants expresses the observation that simple algebraic combina... more The concept of canopy spectral invariants expresses the observation that simple algebraic combinations of leaf and canopy spectral transmittance and reflectance become wavelength independent and determine a small set of canopy structure specific variables. This set includes the canopy interceptance, the recollision and the escape probabilities. These variables specify an accurate relationship between the spectral response of a vegetation canopy to the incident solar radiation at the leaf and the canopy scale and allow for a simple and accurate parameterization for the partitioning of the incoming radiation into canopy transmission, reflection and absorption at any wavelength in the solar spectrum. This paper presents a solid theoretical basis for spectral invariant relationships reported in literature with an emphasis on their accuracies in describing the shortwave radiative properties of the three-dimensional vegetation canopies. The analysis of data on leaf and canopy spectral transmittance and reflectance collected during the international field campaign in Flakaliden, Sweden, June 25 -July 4, 2002 supports the proposed theory. The results presented here are essential to both modeling and remote sensing communities because they allow the separation of the structural and radiometric components of the measured/modeled signal. The canopy spectral invariants offer a simple and accurate parameterization for the shortwave radiation block in many global models of climate, hydrology, biogeochemistry, and ecology. In remote sensing applications, the information content of hyperspectral data can be fully exploited if the wavelength independent variables can be retrieved, for they can be more directly related to structural characteristics of the three dimensional vegetation canopy.

Research paper thumbnail of Reply to Ollinger et al.: Remote sensing of leaf nitrogen and emergent ecosystem properties

Proceedings of the National Academy of Sciences, 2013

Various physical, chemical, and physiological processes, including canopy structure, impact surfa... more Various physical, chemical, and physiological processes, including canopy structure, impact surface reflectance. Remote sensing aims to derive ecosystem properties and their functional relationships, given these impacts. Ollinger et al. (1) do not distinguish between the forward and inverse problems in radiative transfer and, hence, misrepresent our results (2). The authors also suggest our conclusions are based on a subset of data from ref. 3, which is not the case.

Research paper thumbnail of Reply to Townsend et al.: Decoupling contributions from canopy structure and leaf optics is critical for remote sensing leaf biochemistry

Proceedings of the National Academy of Sciences, 2013

Research paper thumbnail of Advances in pattern recognition applications and methods

Neurocomputing, 2014

ABSTRACT Learning large Bayesian networks (BN) from data is a challenging problem due to the vast... more ABSTRACT Learning large Bayesian networks (BN) from data is a challenging problem due to the vastness of the structure space. An effective way to turn this problem affordable is the use of super-structures-SS (undirected graphs that contain the BN skeleton). ...

Research paper thumbnail of Filter-Type Variable Selection Based on Information Measures for Regression Tasks

Entropy, 2012

This paper presents a supervised variable selection method applied to regression problems. This m... more This paper presents a supervised variable selection method applied to regression problems. This method selects the variables applying a hierarchical clustering strategy based on information measures. The proposed technique can be applied to single-output regression datasets, and it is extendable to multi-output datasets. For single-output datasets, the method is compared against three other variable selection methods for regression on four datasets. In the multi-output case, it is compared against other state-of-the-art method and tested using two regression datasets. Two different figures of merit are used (for the single and multi-output cases) in order to analyze and compare the performance of the proposed method.

Research paper thumbnail of Multi-dimensional compressive imaging

Emerging Technologies in Security and Defence; and Quantum Security II; and Unmanned Sensor Systems X, 2013

ABSTRACT In this keynote address paper, we present an overview of our previously published work o... more ABSTRACT In this keynote address paper, we present an overview of our previously published work on using compressive sensing in multi-dimensional imaging. We shall examine a variety of multi dimensional imaging approaches and applications, including 3D multi modal imaging integrated with polarimetric and multi spectral imaging, integral imaging and digital holography. This Keynote Address paper is an overview of our previously reported work on 3D imaging with compressive sensing.

Research paper thumbnail of ICPRAM 2012 - Proceedings of the 1st International Conference on Pattern Recognition Applications and Methods, Volume 1, Vilamoura, Algarve, Portugal, 6-8 February, 2012

Research paper thumbnail of Non-invasive Melanoma Diagnosis using Multispectral Imaging

The early analysis of pigmented skin lesions is important for clinicians in order to recognize ma... more The early analysis of pigmented skin lesions is important for clinicians in order to recognize malignant melanoma. However, it is difficult to differentiate it from benign skin lesions due to their similarity based on their appearance. Since melanoma has a tendency to grow inside the skin and the depth of penetration of light into the skin is wavelength dependent, a multispectral imaging acquisition and processing approach to classify pigmented lesions as melanoma seems appropriate. This paper presents a method to diagnose melanoma lesions over a group of 26 samples acquired with a multispectral system, where 6 of them are melanomas, and the other 20 are other types of pigmented lesions. A Leave-One-Out strategy is used to create the training/test set. The classification imbalance problem inherent to this dataset is alleviated using a SMOT E technique. The random component of the SMOT E methodology is dealt with running it 25 times and a Qualified Majority Voting (QMV ) scheme is used to do the final classification, using SV M. Results show this strategy allows to obtain competitive classification quality results.

Research paper thumbnail of Security authentication using phase-encoded nanoparticle structures and polarized light

Optics Letters, 2015

Phase-encoded nanostructures such as quick response (QR) codes made of metallic nanoparticles are... more Phase-encoded nanostructures such as quick response (QR) codes made of metallic nanoparticles are suggested to be used in security and authentication applications. We present a polarimetric optical method able to authenticate random phase-encoded QR codes. The system is illuminated using polarized light, and the QR code is encoded using a phase-only random mask. Using classification algorithms, it is possible to validate the QR code from the examination of the polarimetric signature of the speckle pattern. We used Kolmogorov-Smirnov statistical test and Support Vector Machine algorithms to authenticate the phase-encoded QR codes using polarimetric signatures.

Research paper thumbnail of Feature selection in regression tasks using conditional mutual information

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2011

This paper presents a supervised feature selection method applied to regression problems. The sel... more This paper presents a supervised feature selection method applied to regression problems. The selection method uses a Dissimilarity matrix originally developed for classification problems, whose applicability is extended here to regression and built using the conditional mutual information between features with respect to a continuous relevant variable that represents the regression function.

Research paper thumbnail of Effect of Denoising in Band Selection for Regression Tasks in Hyperspectral Datasets

IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2013

ABSTRACT This paper presents a comparative analysis of six band selection methods applied to hype... more ABSTRACT This paper presents a comparative analysis of six band selection methods applied to hyperspectral datasets for biophysical variable estimation problems, where the effect of denoising on band selection performance has also been analyzed. In particular, we consider four hyperspectral datasets and three regressors of different nature (ε-SVR, Regression Trees, and Kernel Ridge Regression). Results show that the denoising approach improves the band selection quality of all the tested methods. We show that noise filtering is more beneficial for the selection methods that use an estimator based on the whole dataset for the prediction of the output than for methods that use strategies based on local information (neighboring points).

Research paper thumbnail of Band selection in spectral imaging for classification and regression tasks using information theoretic measures

2011 10th Euro-American Workshop on Information Optics, 2011

In this paper we present three different methodologies of band selection for hyperspectral data s... more In this paper we present three different methodologies of band selection for hyperspectral data sets applied to classification and regression tasks using Information Theory measures. In one of the cases, the bands will be selected having information about the classification labels of the data points (supervised classification). In the second one, no information about the target labels is required (unsupervised classification). In the third problem, the target variables are of continuous nature and are also available (supervised regression).

Research paper thumbnail of Band selection in spectral imaging for non-invasive melanoma diagnosis

Biomedical Optics Express, 2013

A method consisting of the combination of the Synthetic Minority Over-Sampling TEchnique (SMOTE) ... more A method consisting of the combination of the Synthetic Minority Over-Sampling TEchnique (SMOTE) and the Sequential Forward Floating Selection (SFFS) technique is used to do band selection in a highly imbalanced, small size, two-class multispectral dataset of melanoma and non-melanoma lesions. The aim is to improve classification rate and help to identify those spectral bands that have a more important role in melanoma detection. All the processing steps were designed taking into account the low number of samples in the dataset, situation that is quite common in medical cases. The training/test sets are built using a Leave-One-Out strategy. SMOTE is applied in order to deal with the imbalance problem, together with the Qualified Majority Voting scheme (QMV). Support Vector Machines (SVM) is the classification method applied over each balanced set. Results indicate that all melanoma lesions are correctly classified, using a low number of bands, reaching 100% sensitivity and 72% specificity when considering nine (out of a total of 55) spectral bands.

Research paper thumbnail of One-class classification techniques in image recognition problems

2013 12th Workshop on Information Optics (WIO), 2013

ABSTRACT The one class classification problem is different from the binary/multi-class classifica... more ABSTRACT The one class classification problem is different from the binary/multi-class classification problem in the sense that the negative class is not present or not properly character-ized. The problem of classifying positive (target) cases in the absence of appropriately defined negative cases (outliers) has been given an increased attention during the last few years. In this paper we give a brief overview about the methods used for one class classification applied in three fields of interest: (a) remote sensing, (b) 3D image retrieval and modeling and (c) medical image/data analysis/processing. We also present results of the application of one of the methods (one-class SYM) for the detection of vegetated areas in a hyperspectral remote sensing image.

Research paper thumbnail of PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES

Research paper thumbnail of Canopy spectral invariants for remote sensing and model applications

Remote Sensing of Environment, 2007

The concept of canopy spectral invariants expresses the observation that simple algebraic combina... more The concept of canopy spectral invariants expresses the observation that simple algebraic combinations of leaf and canopy spectral transmittance and reflectance become wavelength independent and determine a small set of canopy structure specific variables. This set includes the canopy interceptance, the recollision and the escape probabilities. These variables specify an accurate relationship between the spectral response of a vegetation canopy to the incident solar radiation at the leaf and the canopy scale and allow for a simple and accurate parameterization for the partitioning of the incoming radiation into canopy transmission, reflection and absorption at any wavelength in the solar spectrum. This paper presents a solid theoretical basis for spectral invariant relationships reported in literature with an emphasis on their accuracies in describing the shortwave radiative properties of the three-dimensional vegetation canopies. The analysis of data on leaf and canopy spectral transmittance and reflectance collected during the international field campaign in Flakaliden, Sweden, June 25 -July 4, 2002 supports the proposed theory. The results presented here are essential to both modeling and remote sensing communities because they allow the separation of the structural and radiometric components of the measured/modeled signal. The canopy spectral invariants offer a simple and accurate parameterization for the shortwave radiation block in many global models of climate, hydrology, biogeochemistry, and ecology. In remote sensing applications, the information content of hyperspectral data can be fully exploited if the wavelength independent variables can be retrieved, for they can be more directly related to structural characteristics of the three dimensional vegetation canopy.

Research paper thumbnail of Reply to Ollinger et al.: Remote sensing of leaf nitrogen and emergent ecosystem properties

Proceedings of the National Academy of Sciences, 2013

Various physical, chemical, and physiological processes, including canopy structure, impact surfa... more Various physical, chemical, and physiological processes, including canopy structure, impact surface reflectance. Remote sensing aims to derive ecosystem properties and their functional relationships, given these impacts. Ollinger et al. (1) do not distinguish between the forward and inverse problems in radiative transfer and, hence, misrepresent our results (2). The authors also suggest our conclusions are based on a subset of data from ref. 3, which is not the case.

Research paper thumbnail of Reply to Townsend et al.: Decoupling contributions from canopy structure and leaf optics is critical for remote sensing leaf biochemistry

Proceedings of the National Academy of Sciences, 2013

Research paper thumbnail of Advances in pattern recognition applications and methods

Neurocomputing, 2014

ABSTRACT Learning large Bayesian networks (BN) from data is a challenging problem due to the vast... more ABSTRACT Learning large Bayesian networks (BN) from data is a challenging problem due to the vastness of the structure space. An effective way to turn this problem affordable is the use of super-structures-SS (undirected graphs that contain the BN skeleton). ...

Research paper thumbnail of Filter-Type Variable Selection Based on Information Measures for Regression Tasks

Entropy, 2012

This paper presents a supervised variable selection method applied to regression problems. This m... more This paper presents a supervised variable selection method applied to regression problems. This method selects the variables applying a hierarchical clustering strategy based on information measures. The proposed technique can be applied to single-output regression datasets, and it is extendable to multi-output datasets. For single-output datasets, the method is compared against three other variable selection methods for regression on four datasets. In the multi-output case, it is compared against other state-of-the-art method and tested using two regression datasets. Two different figures of merit are used (for the single and multi-output cases) in order to analyze and compare the performance of the proposed method.

Research paper thumbnail of Multi-dimensional compressive imaging

Emerging Technologies in Security and Defence; and Quantum Security II; and Unmanned Sensor Systems X, 2013

ABSTRACT In this keynote address paper, we present an overview of our previously published work o... more ABSTRACT In this keynote address paper, we present an overview of our previously published work on using compressive sensing in multi-dimensional imaging. We shall examine a variety of multi dimensional imaging approaches and applications, including 3D multi modal imaging integrated with polarimetric and multi spectral imaging, integral imaging and digital holography. This Keynote Address paper is an overview of our previously reported work on 3D imaging with compressive sensing.