Jaime R. Ticay-Rivas

Image Processing for Pollen Classification

Automatic recognition of pollen grains can overcome these problems, producing purely objective results faster. Such a tool would provide invaluable in the studies of flora. This advantages were obvious for Flenley [5] [6], who proposed the implementation of an automatic pollen grain classification system in 1968. However, the idea was intractable at that time. Mainly, because of technology restrictions. Nowadays, technology is not a barrier any more, and the discussed system is a reality thanks to computer vision.

Electroencephalogram subject identification: A review

This is, to the best of the authors knowledge, the first complete research on the state of the art on EEG based subject identification. As well as covering the full story of this field (from 1980 to 2013), an overview of the findings made in genetic and neurophysiology areas, from which it is based, is also provided. After a comprehensive search, 109 biometric publications were found and studied, from which 88 were finally included in this document. A categorization of papers is proposed based on the recording paradigm. The most used databases, some of them public, have been identified and named to allow the comparison of results from these and future works. The findings of this work show that, although basic questions remain to be answered, the EEG, and specially its power spectrum in the range of the alpha rhythm, contains subject specific information that can be used for classification. Moreover, approaches such as a multi-day-session training, the fusion of information from different electrodes and bands, and Support Vector Machines are recommended to maximize the systems performance. All in all, the problem of subject identification by means of their EEG is harder than initially expected, as it relies on information extracted from complex heterogeneous EEG traits which are the results of elaborated models of inheritance, which in turn makes the problem very sensitive to its variables (time, frequency, space, recording paradigm and algorithms).

Pollen classification based on contour features

Conserving earths biodiversity for future generations is a fundamental global task, where automated recognition of pollen species by means of computer vision represents a highly prioritized issue. This work focuses on analysis and classification stages. The morphological details of the contour are proposed as pollen grains discriminative features. The approach has been developed as a robust pollen identification based on an HMM kernel. A Vector Support Machine was used as classifier. The principal contribution in this work, in terms of the use of the HMM is the gradient optimisation problem implementation in the SVM. 47 tropical honey plant species have been classified achieving a mean of 93.8% ± 1.43 of success.

Hand shape identification on multirange images

Abstract A hand-shape based biometric identification system which is independent of the image spectrum range is proposed here. Two different spectrum ranges; visible and mid-range infrared, were used to validated the architecture, which maintained the accuracy and stability levels between ranges. In particular, three public databases were tested, obtaining accuracies over 99.9% using a 40% hold-out cross-validation approach. Discrete Hidden Markov Models (DHMM) representing each target identification class was trained with angular chain descriptors. A kernel was then extracted from the trained DHMM and applied as a feature extraction method. Finally, supervised Support Vector Machines were used to classify the extracted features.

Features extraction techniques for pollen grain classification

Abstract An extensive study on pollen grain identification is presented in this work. A combination of geometrical and texture characteristics is proposed as pollen grain discriminative features as well as the usage of the most popular feature extraction techniques. Multi-Layer Neural Network and Least Square Support Vector Machines (LS-SVM) with Radial Basis Function were used as classifier systems. K -fold and hold-out cross-validation techniques were applied in order to achieve reliable results. When testing with a 17-species database, the combination of the proposed set of features processed by Linear Discriminant Analysis and the LS-SVM has provided the best performance, reaching a 94.92%±0.61 of success rate. Subsequently, the combination of both classifier methods provided better results, achieving 95.27%±0.49 of accuracy.

Spider specie identification and verification based on pattern recognition of it cobweb

Biodiversity conservation is a global priority where the study of every type of living form is a fundamental task. Inside the huge number of the planet species, spiders play an important role in almost every habitat. This paper presents a comprehensive study on the reliability of the most used features extractors to face the problem of spider specie recognition by using their cobwebs, both in identification and verification modes. We have applied a preprocessing to the cobwebs images in order to obtain only the valid information and compute the optimal size to reach the highest performance. We have used the principal component analysis (PCA), independent component analysis (ICA), Discrete Cosine Transform (DCT), Wavelet Transform (DWT) and discriminative common vectors as features extractors, and proposed the fusion of several of them to improve the systems performance. Finally, we have used the Least Square Vector Support Machine with radial basis function as a classifier. We have implemented K-Fold and Hold-Out cross-validation techniques in order to obtain reliable results. PCA provided the best performance, reaching a 99.65%+/-0.21 of success rate in identification mode and 99.98%+/-0.04 of the area under de Reveicer Operating Characteristic (ROC) curve in verification mode. The best combination of features extractors was PCA, DCT, DWT and ICA, which achieved a 99.96%+/-0.16 of success rate in identification mode and perfect verification.

Pollen Classification Based on Geometrical, Descriptors and Colour Features Using Decorrelation Stretching Method

Saving earth’s biodiversity for future generations is an important global task, where automatic recognition of pollen species by means of computer vision represents a highly prioritized issue. This work focuses on analysis and classification stages. A combination of geometrical measures, Fourier descriptors of morphological details using Discrete Cosine Transform (DCT) in order to select their most significant values, and colour information over decorrelated stretched images are proposed as pollen grains discriminative features. A Multi-Layer neural network was used as classifier applying scores fusion techniques. 17 tropical honey plant species have been classified achieving a mean of 96.49% ± 1.16 of success.

Spider recognition by biometric web analysis

Saving earths biodiversity for future generations is an important global task. Spiders are creatures with a fascinating behaviour, overall in the way they build their webs. This is the reason this work proposed a novel problem: the used of spider webs as a source of information for specie recognition. To do so, biometric techniques such as image processing tools, Principal Component Analysis, and Support Vector Machine have been used to build a spider web identification system. With a database built of images from spider webs of three species, the system reached a best performance of 95,44 % on a 10 K-Folds crossvalidation procedure.

Real cepstrums on electroencephalogram biometric identification

In a time when personal data circulates constantly between devices and within the cloud, biometric security systems represents one of the most viable security solutions. A relatively new biometric modality based on the individuals Electroencephalogram (EEG) is starting now to gain popularity among researchers. Its relevance relay mainly on its prospects of high security and robustness against intruders and the proliferation of consumer EEG devices. In this work we propose the use of real cepstrums as descriptors of the subject traits within the EEG. When evaluated with each of the 14 conditions of the 100-subjects BCI2000 database, the proposed approach achieved classification accuracies between 86.88% and 94.91% using only the first 5% of the computed cepstral coefficients (13 coefficients).

Review: Electroencephalogram subject identification: A review

This is, to the best of the authors knowledge, the first complete research on the state of the art on EEG based subject identification. As well as covering the full story of this field (from 1980 to 2013), an overview of the findings made in genetic and neurophysiology areas, from which it is based, is also provided. After a comprehensive search, 109 biometric publications were found and studied, from which 88 were finally included in this document. A categorization of papers is proposed based on the recording paradigm. The most used databases, some of them public, have been identified and named to allow the comparison of results from these and future works. The findings of this work show that, although basic questions remain to be answered, the EEG, and specially its power spectrum in the range of the alpha rhythm, contains subject specific information that can be used for classification. Moreover, approaches such as a multi-day-session training, the fusion of information from different electrodes and bands, and Support Vector Machines are recommended to maximize the systems performance. All in all, the problem of subject identification by means of their EEG is harder than initially expected, as it relies on information extracted from complex heterogeneous EEG traits which are the results of elaborated models of inheritance, which in turn makes the problem very sensitive to its variables (time, frequency, space, recording paradigm and algorithms).