Manuel Alejandro Ospina Alarcón

Dr. Manuel Alejandro Ospina Alarcón

System Engineering, Universidad de Cartagena, Cartagena, Bolívar, Colombia
orcid

Research Focus & Metrics

Academic Career Length

9Years (2018-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
14 July 2023 in Atmosphere
Verified
Journal Article
Open access
PDF file
GGGabriel E. Chanchí Golondrino
Gabriel E. Chanchí Golondrino
Manuel Alejandro Ospina AlarcónManuel Alejandro Ospina Alarcón
Manuel Alejandro Ospina Alarcón
MSManuel Saba
Manuel Saba
Distance/correlation metrics have emerged as a robust and simplified tool for assessing the spectral characteristics of hyperspectral image pixels and effectively categorizing vegetation within a specific study area. Correlation methods provide a readily deployable and computationally efficient approach, rendering them particularly advantageous for applications in developing nations or regions with limited resources. This article presents a comparative investigation of correlation/distance metrics for the identification of vegetation pixels in hyperspectral imagery. The study facilitates a comprehensive evaluation of five distance and/or correlation metrics, namely, direct correlation, cosine similarity, normalized Euclidean distance, Bray–Curtis distance, and Pearson correlation. Direct correlation and Pearson correlation emerged as the two metrics that demonstrated the highest accuracy in vegetation pixel identification. Using the selected methodologies, a vegetation detection algorithm was implemented and validated using a hyperspectral image of the Manga neighborhood in Cartagena de Indias, Colombia. The spectral library facilitated image processing, while the mathematical calculation of correlations was performed using the numpy and scipy libraries in the Python programming language. Both the approach adopted in this study and the implemented algorithm aim to serve as a point of reference for conducting detection studies on diverse material types in hyperspectral imagery using open-access programming platforms.

18 July 2018 in Computers & Chemical Engineering
Verified
Journal Article
Closed access
CZC. Zuluaga-Bedoya
C. Zuluaga-Bedoya
MRM. Ruiz-Botero
M. Ruiz-Botero
Manuel Alejandro Ospina AlarcónManuel Alejandro Ospina Alarcón
Manuel Alejandro Ospina Alarcón
JGJ. Garcia-Tirado
J. Garcia-Tirado

19 September 2024 in Applied Sciences
Verified
Journal Article
Open access
PDF file
MSManuel Saba
Manuel Saba
DMDavid Enrique Valdelamar Martínez
David Enrique Valdelamar Martínez
LGLeydy K. Torres Gil
Leydy K. Torres Gil
GGGabriel E. Chanchí Golondrino
Gabriel E. Chanchí Golondrino
Manuel Alejandro Ospina AlarcónManuel Alejandro Ospina Alarcón
Manuel Alejandro Ospina Alarcón
This study introduces an innovative approach to evaluate the condition of asbestos–cement (AC) roofs by integrating field data with five distinct supervised learning models: decision trees, KNN, logistic regression, support vector machine, and random forest. A novel methodology for assessing the importance of 380 reflectance bands was employed, offering fresh insights into the key indicators of AC roof deterioration. The research systematically organized and prioritized reflectance bands based on their information gain, optimizing both the selection of relevant bands and the performance of the models in differentiating between low and high intervention priority (LIP and HIP) roofs. The decision tree model, when applied to the top 10 most relevant bands, achieved the highest cross-validation accuracy of 76.047%, making it the most effective tool for identifying AC roof conditions. Additionally, the random forest model demonstrated strong performance across various band groups, further validating its utility. Utilizing the open-source software Weka (version 3.8.6), this study adeptly executed relevance evaluation and model implementation, providing a practical and scalable solution for material characterization, especially in regions where resources for spectral and hyperspectral image analysis are limited. The findings of this study offer valuable tools for government and environmental authorities, particularly in developing countries, where efficient and cost-effective AC roof assessment is crucial for public health and safety. The methodology is adaptable to different urban environments and climatic conditions, supporting global efforts in asbestos management, especially in countries where asbestos regulations are newly implemented. Organized within the CRISP-DM framework, this paper details the methodological phases, presents compelling results on reflectance band relevance, evaluates machine learning models, and concludes with prospects for future research aimed at enhancing asbestos detection and removal strategies.

1 May 2022 in Revista Científica
Verified
Journal Article
Open access
PDF file
GGGabriel E. Chanchí Golondrino
Gabriel E. Chanchí Golondrino
Manuel Alejandro Ospina AlarcónManuel Alejandro Ospina Alarcón
Manuel Alejandro Ospina Alarcón
MSManuel Saba
Manuel Saba
A partir de la creciente tendencia de la agricultura urbana, este trabajo tiene por objetivo la construcción de un sistema IoT para la monitorización y el análisis de variables climatológicas de interés en cultivos de agricultura urbana. La metodología considerada para el desarrollo de la presente investigación está constituida por cuatro fases: selección de herramientas y tecnologías, diseño de arquitectura del sistema IoT, construcción de prototipo del sistema y estudio de caso en el contexto del cultivo de lechuga. Como resultado de la presente investigación, se construyó un sistema IoT basado en herramientas de hardware y software libre, el cual está articulado dentro de la arquitectura convencional de cuatro capas de IoT (captura, almacenamiento, análisis y visualización). Con respecto a las soluciones existentes, la ventaja del sistema es el uso de plataformas portables SBC, así como la inclusión de modelos de aprendizaje automático dentro de la capa de análisis. A partir del estudio de caso, desarrollado sobre un cultivo de lechuga casero, se concluye que las herramientas escogidas permiten realizar de manera adecuada la captura, la monitorización y el análisis de variables climatológicas de interés en cultivos de agricultura urbana. Del mismo modo, se concluye que los modelos de análisis considerados pueden ser personalizados teniendo en cuenta las características agroclimáticas de cada cultivo y que resultan útiles para la toma de decisiones relacionadas con la fisiología de los cultivos.

22 August 2024 in INGENIERÍA Y COMPETITIVIDAD
Verified
Journal Article
Open access
PDF file
GGGabriel E. Chanchí Golondrino
Gabriel E. Chanchí Golondrino
Manuel Alejandro Ospina AlarcónManuel Alejandro Ospina Alarcón
Manuel Alejandro Ospina Alarcón
MSManuel Saba
Manuel Saba
Introduction: hyperspectral images, unlike conventional images, are composed of numerous channels that provide detailed information about the spectral signatures of objects. This allows for the identification of the materials that make them up, and given their potential for detecting environmental changes, identifying vegetation in urban settings using effective computational methods becomes relevant.Objective: the objective of this research is to propose a computational method based on Fourier analysis for detecting vegetation in hyperspectral images.Methods: the research was developed in four methodological phases: selection of technologies, acquisition of the characteristic vegetation pixel, determination of phase similarity between the characteristic pixel and vegetation and non-vegetation pixels, validation of the method on a test hyperspectral image. A method was implemented using the spectral and numpy libraries in Python.Results: the Fourier analysis yielded an average phase similarity of 89.89% and a minimum similarity of 64.54% between the characteristic vegetation pixel and 100 training vegetation pixels. For non-vegetation pixels, the average phase similarity was 42.19%, with a maximum similarity of 63.98%. These results indicate that the proposed method successfully differentiates between vegetation and non-vegetation pixels.Conclusion: the results demonstrate that the Fourier-based method can accurately identify vegetation areas in hyperspectral images, showing non-overlapping phase similarities between vegetation and non-vegetation. This validates the effectiveness of the proposed approach in detecting vegetation in urban environments.

30 December 2022 in INGENIERÍA Y COMPETITIVIDAD
Verified
Journal Article
Open access
PDF file
GGGabriel E. Chanchí Golondrino
Gabriel E. Chanchí Golondrino
Manuel Alejandro Ospina AlarcónManuel Alejandro Ospina Alarcón
Manuel Alejandro Ospina Alarcón
LSLuis Freddy Muñoz Sanabria
Luis Freddy Muñoz Sanabria
Taking into account the wide diffusion that data analytics has had in different application areas and considering the scarcity of specific datasets associated with cybercrime within open data strategies in Colombia, this article aims to characterize cybercrime in the department of Cundinamarca, through the use of exploratory analysis and machine learning techniques. The present research was developed through 4 methodological phases: data adequacy, exploratory data analysis, application of machine learning models and finally generation of value-added information. For the development of the proposed study, a dataset was formed from the dataset of 35,000 records published by the National Police in the open data portal of Colombia, which addresses high-impact crimes within the department of Cundinamarca and occurred during the first half of 2021. The cybercrime dataset has a total of 1513 records and includes attributes such as: day, quarter, municipality, area, victim, age and crime, so that at the exploratory analysis level, descriptive statistics methods were applied on the different attributes, while at the machine learning level, the association rules and clustering models were applied in order to determine respectively the relationship of the attributes with the type of crime, and the representative groups formed by relating the age with the type of crime and the municipality with the type of crime. The study developed allowed to demonstrate the usefulness and potential of data analytics techniques in the field of cybersecurity, in order to support decision making by the relevant authorities. Teniendo en cuenta la amplia difusión que ha tenido la analítica de datos en diferentes ámbitos de aplicación y considerando los escases de datasets específicos asociados a los delitos informáticos dentro de las estrategias de datos abiertos en Colombia, este artículo tiene como objetivo realizar la caracterización de los delitos informáticos del departamento de Cundinamarca, mediante el uso de técnicas de análisis exploratorio y machine learning. La presente investigación fue desarrollada mediante 4 fases metodológicas: adecuación de los datos, análisis exploratorio de los datos, aplicación de modelos de machine learning y finalmente generación de información de valor agregado. Para el desarrollo del estudio propuesto, se conformó un conjunto de datos a partir del dataset de 35000 registros publicado por la Policía Nacional en el portal de datos abiertos de Colombia, el cual aborda los delitos de alto impacto dentro del departamento de Cundinamarca y ocurridos durante el primer semestre de 2021. El dataset de delitos cibernéticos conformado cuenta con un total de 1513 registros e incluye atributos tales como: día, trimestre, municipio, zona, victima, edad y delito, de tal modo que a nivel del análisis exploratorio se aplicaron métodos de estadística descriptiva sobre los diferentes atributos, mientras que a nivel de machine learning se hizo uso de los modelos de reglas de asociación y clustering con el fin de determinar de manera respectiva la relación de los atributos con el tipo de delito, y los grupos representativos que se forman al relacionar la edad con el tipo de delito y el municipio con el tipo de delito. El estudio desarrollado permitió demostrar la utilidad y potencialidad que tienen las técnicas de analítica de datos en el campo de la ciberseguridad, de cara a apoyar la toma de decisiones por parte de las autoridades pertinentes.