Author: MIGUEL ANGEL MEDINA PEREZ

Algoritmos basados en tríos de minucias para la verificación e identificación de huellas dactilares

MIGUEL ANGEL MEDINA PEREZ (2014)

El reconocimiento automatizado de huellas dactilares y palmares es un área activa de investigación en la actualidad. Las huellas dactilares (palmares) se forman a partir del contacto de los dedos (las palmas) de las manos con las superficies de los objetos. Dentro de los sistemas automatizados de reconocimientos de huellas, los algoritmos de comparación de huellas constituyen un componente de vital importancia. Un algoritmo de comparación de huellas es una secuencia finita de pasos lógicos, que compara las huellas descritas por rasgos (o abstracciones) y retorna un valor de similitud. Uno de los enfoques seguidos en la literatura para representar y comparar las huellas es el uso de tríos de minucias. Las minucias son los puntos donde la continuidad de las crestas papilares (glándulas de secreción de sudor en dedos, palmas de la mano y pies) se rompe. En esta tesis se hace un análisis de los algoritmos basados en tríos de minucias y se identifican las limitaciones que degradan su eficacia. Como resultado de la investigación, se propone una nueva forma de representación y función de comparación de tríos de minucias que supera las limitaciones identificadas en la literatura. Basado en la nueva forma de representación y comparación de tríos de minucias, se construye un algoritmo de comparación para la verificación de huellas dactilares; otro para la verificación de huellas palmares y un tercero para la identificación de huellas latentes (dactilares y palmares). El objetivo de las aplicaciones de verificación de huellas consiste esencialmente en determinar dadas dos impresiones (huellas formadas voluntariamente en condiciones controladas), si estas proceden del mismo dedo. Por otra parte, el objetivo de las aplicaciones de identificación de huellas latentes (huella formadas involuntariamente en condiciones no controladas) consiste básicamente en, dada una huella latente, buscar todas las huellas procedentes del mismo dedo (o palma) presentes en una base de datos. Los algoritmos propuestos en esta investigación usan los tríos de minucias para hallar múltiples conjuntos de pares de minucias coincidentes. Los conjuntos obtenidos son mezclados de acuerdo a una novedosa heurística para obtener la mayor cantidad de coincidencias entre las huellas que se comparan.

The automated fingerprint recognition is currently an active research area. Fingerprints (palmprints) are formed from the contact of the fingers (palms) of the hands with the surfaces of objects. The fingerprint matching algorithms are important components of automated fingerprint recognition systems. A fingerprint (palmprint) matching algorithm is a finite sequence of logical steps, which compares the fingerprints (palmprints) described by features (or abs- tractions) and returns a similarity value. An approach followed in the literature for fingerprint matching and representation is the use of minutiae triplets. Minutiae are the points where the continuity of ridges (sweat glands in fingers, hand plants, and soles) break. In this thesis, the fingerprint matching algorithms based on minutiae triplets are analyzed and the drawbacks that affect their accuracy are identified. As a result of this research, it is proposed a feature representation and a matching function for minutiae triplets that overcome the identified limitations in the literature. Based on the new representation and matching function for minutiae triplets, a matching algorithm is created for fingerprint verification, another matching algorithm is created for palmprint verification, and a third matching algorithm is created for latent fingerprint and palmprint identification. The aim of finger- print (palmprint) verification applications is essentially determining whether two impressions (prints formed voluntarily in controlled conditions) comes from the same finger (palm) or not. In the other hand, the aim of latent print identification applications is essentially, given a latent print (prints formed involuntarily in uncontrolled conditions), searching all the prints in a database that comes from the same finger (or palm). The proposed algorithms use minutiae triplets in order to find multiple sets of matching minutiae. These sets are merged according to a new heuristic in order to obtain the highest amount of matching minutiae between the compared prints.

Doctoral thesis

RECONOCIMIENTO DE PATRONES BIOMETRIA RECONOCIMIENTO DE HUELLAS CIENCIAS FÍSICO MATEMÁTICAS Y CIENCIAS DE LA TIERRA MATEMÁTICAS CIENCIA DE LOS ORDENADORES CIENCIA DE LOS ORDENADORES

LCMine: An efficient algorithm for mining discriminative regularities and its application in supervised classification

MILTON GARCÍA BORROTO JOSE FRANCISCO MARTINEZ TRINIDAD JESUS ARIEL CARRASCO OCHOA MIGUEL ANGEL MEDINA PEREZ (2010)

In this paper, we introduce an efficient algorithm for mining discriminative regularities on databases with mixed and incomplete data. Unlike previous methods, our algorithm does not apply an a priori discretization on numerical features; it extracts regularities from a set of diverse decision trees, induced with a special procedure. Experimental results show that a classifier based on the regularities obtained by our algorithm attains higher classification accuracy, using fewer discriminative regularities than those obtained by previous pattern-based classifiers. Additionally, we show that our classifier is competitive with traditional and state-of-the-art classifiers.

Article

Discriminative regularities Emerging patterns Mixed incomplete data Comprehensible classifiers CIENCIAS FÍSICO MATEMÁTICAS Y CIENCIAS DE LA TIERRA MATEMÁTICAS CIENCIA DE LOS ORDENADORES CIENCIA DE LOS ORDENADORES

Improving fingerprint verification using minutiae triplets

MIGUEL ANGEL MEDINA PEREZ MILTON GARCÍA BORROTO ANDRES EDUARDO GUTIERREZ RODRIGUEZ Leopoldo Altamirano Robles (2012)

Improving fingerprint matching algorithms is an active and important research area in fingerprint recognition. Algorithms based on minutia triplets, an important matcher family, present some drawbacks that impact their accuracy, such as dependency to the order of minutiae in the feature, insensitivity to the reflection of minutiae triplets, and insensitivity to the directions of the minutiae relative to the sides of the triangle. To alleviate these drawbacks, we introduce in this paper a novel fingerprint matching algorithm, named M3gl. This algorithm contains three components: a new feature representation containing clockwise-arranged minutiae without a central minutia, a new similarity measure that shifts the triplets to find the best minutiae correspondence, and a global matching procedure that selects the alignment by maximizing the amount of global matching minutiae. To make M3gl faster, it includes some optimizations to discard non-matching minutia triplets without comparing the whole representation. In comparison with six verification algorithms, M3gl achieves the highest accuracy in the lowest matching time, using FVC2002 and FVC2004 databases.

Article

Fingerprint verification Minutiae descriptor Minutiae triplet CIENCIAS FÍSICO MATEMÁTICAS Y CIENCIAS DE LA TIERRA MATEMÁTICAS CIENCIA DE LOS ORDENADORES CIENCIA DE LOS ORDENADORES

Image annotation as Text-Image matching: Challenge design and results

Luis Luis Pellegrin OCTAVIO LOYOLA GONZALEZ JOSE ORTIZ BEJAR MIGUEL ANGEL MEDINA PEREZ ANDRES EDUARDO GUTIERREZ RODRIGUEZ Eric Sadit Téllez Avila MARIO GRAFF GUERRERO SABINO MIRANDA JIMENEZ Daniela Moctezuma MAURICIO ALFONSO GARCIA LIMON ALICIA MORALES REYES CARLOS ALBERTO REYES GARCIA Eduardo Morales Manzanares Hugo Jair Escalante (2019)

This paper describes the design of the 2017 RedICA: Text-Image Matching (RICATIM) challenge, including the dataset generation, a complete analysis of results, and the descriptions of the top-ranked developed methods. The academic challenge explores the feasibility of a novel binary image classification scenario, where each instance corresponds to the concatenation of learned representations of an image and a word. Instances are labeled as positive if the word is relevant for describing the visual content of the image, and negative otherwise. This novel approach of the image classification problem poses an alternative scenario where any text-image pair can be represented in such space, so any word could be considered for describing an image. The proposed methods are diverse and competitive, showing considerable improvements over the proposed baselines.

Article

Text-image matching Image annotation Multimodal information processing INGENIERÍA Y TECNOLOGÍA CIENCIAS TECNOLÓGICAS OTRAS ESPECIALIDADES TECNOLÓGICAS OTRAS OTRAS

Isolation and characterization of mercury resistant trichoderma strains from soil with high levels of mercury and its effects on Arabidopsis thaliana mercury uptake

JOSE LUIS HERNANDEZ FLORES JOSE GERARDO BERNAL MELO ANDRES CRUZ HERNANDEZ MIGUEL ANGEL RAMOS LOPEZ CARLOS SALDAÑA GUTIERREZ SERGIO DE JESUS ROMERO GOMEZ VICTOR PEREZ MORENO ROSA PAULINA CALVILLO MEDINA JOSE HUMBERTO VALENZUELA SOTO Sergio Pacheco Hernández George H. Jones J Sergio Casas Flores JUAN CAMPOS GUILLEN (2018)

"Traditional mining activities are usually correlated with high levels of soil pollution, which is a major environmental concern. Extensive mining activities have taken place in the San Joaquin region in the State of Querétaro, México resulting in high levels of mercury soil pollution (up to 1532 ± 300 mg/kg). We isolated mercury-resistant fungal strains from the San Joaquin region soils and identified them through morphologic characteristics and ITS rDNA region sequence analysis. We determined that fungi isolated belong to the genus Trichoderma. All the isolates selected showed the ability to catalyze the volatilization of Hg. For air sampling, an active sampling device was constructed and using acid KMnO4 as an absorbent, the concentration of mercury in solution was determined through the cold vapor atomic absorption method. The results show mercury volatilization from the fungal species assay, with a maximum of 213.04 ± 32.6 µg/m3 while mycelium accumulation ranged from less than 17.5 ± 2.9 to 20.0 ± 3.4 µg/g. The fungal isolates were also evaluated for their ability to reduce mercury uptake in Arabidopsis thaliana. These observations suggest the utility of Trichoderma for the mobilization of mercury in those contaminated soils."

Article

Volatilization Mercury Trichoderma Soil Bioremediation BIOLOGÍA Y QUÍMICA CIENCIAS DE LA VIDA MICROBIOLOGÍA MICROBIOLOGÍA