Examinando por Materia "remote sensing"
Mostrando 1 - 8 de 8
Resultados por página
Opciones de ordenación
Ítem Exploring the Potential of Machine Learning for Automatic Slum Identification from VHR Imagery(MDPI AG, 2017-09-01) Duque JC; Patiño, Jorge; Betancourt, Alejandro; Universidad EAFIT. Departamento de Economía y Finanzas; Research in Spatial Economics (RISE)Slum identification in urban settlements is a crucial step in the process of formulation of pro-poor policies. However, the use of conventional methods for slum detection such as field surveys can be time-consuming and costly. This paper explores the possibility of implementing a low-cost standardized method for slum detection. We use spectral, texture and structural features extracted from very high spatial resolution imagery as input data and evaluate the capability of three machine learning algorithms (Logistic Regression, Support Vector Machine and Random Forest) to classify urban areas as slum or no-slum. Using data from Buenos Aires (Argentina), Medellin (Colombia) and Recife (Brazil), we found that Support Vector Machine with radial basis kernel delivers the best performance (with F2-scores over 0.81). We also found that singularities within cities preclude the use of a unified classification model.Ítem Hygroscopic growth study in the framework of EARLINET during the SLOPE i campaign: Synergy of remote sensing and in situ instrumentation(Copernicus GmbH, 2018-05-18) Bedoya-Velásquez A.E.; Navas-Guzmán F.; Granados-Muñoz M.J.; Titos G.; Román R.; Andrés Casquero-Vera J.; Ortiz-Amezcua P.; Antonio Benavent-Oltra J.; De Arruda Moreira G.; Montilla-Rosero E.; Hoyos C.D.; Artiñano B.; Coz E.; Olmo-Reyes F.J.; Alados-Arboledas L.; Guerrero-Rascado J.L.; Universidad EAFIT. Departamento de Ciencias Básicas; Óptica AplicadaThis study focuses on the analysis of aerosol hygroscopic growth during the Sierra Nevada Lidar AerOsol Profiling Experiment (SLOPE I) campaign by using the synergy of active and passive remote sensors at the ACTRIS Granada station and in situ instrumentation at a mountain station (Sierra Nevada, SNS). To this end, a methodology based on simultaneous measurements of aerosol profiles from an EARLINET multi-wavelength Raman lidar (RL) and relative humidity (RH) profiles obtained from a multi-instrumental approach is used. This approach is based on the combination of calibrated water vapor mixing ratio (r) profiles from RL and continuous temperature profiles from a microwave radiometer (MWR) for obtaining RH profiles with a reasonable vertical and temporal resolution. This methodology is validated against the traditional one that uses RH from co-located radiosounding (RS) measurements, obtaining differences in the hygroscopic growth parameter (?) lower than 5% between the methodology based on RS and the one presented here. Additionally, during the SLOPE I campaign the remote sensing methodology used for aerosol hygroscopic growth studies has been checked against Mie calculations of aerosol hygroscopic growth using in situ measurements of particle number size distribution and submicron chemical composition measured at SNS. The hygroscopic case observed during SLOPE I showed an increase in the particle backscatter coefficient at 355 and 532nm with relative humidity (RH ranged between 78 and 98%), but also a decrease in the backscatter-related Ångström exponent (AE) and particle linear depolarization ratio (PLDR), indicating that the particles became larger and more spherical due to hygroscopic processes. Vertical and horizontal wind analysis is performed by means of a co-located Doppler lidar system, in order to evaluate the horizontal and vertical dynamics of the air masses. Finally, the Hänel parameterization is applied to experimental data for both stations, and we found good agreement on ? measured with remote sensing (?532 0.48 ± 0.01 and ?355 0.40 ± 0.01) with respect to the values calculated using Mie theory (?532 0.53 ± 0.02 and ?355 0.45 ± 0.02), with relative differences between measurements and simulations lower than 9% at 532nm and 11% at 355nm. © Author(s) 2018.Ítem Inteligencia artificial para detectar la roya en el café(Universidad EAFIT, 2020-12-01) Martinez Guerrero, Christian Alexander; Martinez-Guerrero, Christian Alexander; Velásquez, David; Sánchez, Alejandro; Sarmiento, Sebastian; Toro, Mauricio; Maiza, Mikel; Sierra, Basilio; GIDITIC; Estudios en MantenimientoÍtem Remote sensing-based measurement of Living Environment Deprivation: Improving classical approaches with machine learning(Public Library of Science, 2017-05-02) Arribas-Bel D; Patino JE; Duque JC; Universidad EAFIT. Departamento de Economía y Finanzas; Research in Spatial Economics (RISE)This paper provides evidence on the usefulness of very high spatial resolution (VHR) imagery in gathering socioeconomic information in urban settlements. We use land cover, spectral, structure and texture features extracted from a Google Earth image of Liverpool (UK) to evaluate their potential to predict Living Environment Deprivation at a small statistical area level. We also contribute to the methodological literature on the estimation of socioeconomic indices with remote-sensing data by introducing elements from modern machine learning. In addition to classical approaches such as Ordinary Least Squares (OLS) regression and a spatial lag model, we explore the potential of the Gradient Boost Regressor and Random Forests to improve predictive performance and accuracy. In addition to novel predicting methods, we also introduce tools for model interpretation and evaluation such as feature importance and partial dependence plots, or cross-validation. Our results show that Random Forest proved to be the best model with an R2 of around 0.54, followed by Gradient Boost Regressor with 0.5. Both the spatial lag model and the OLS fall behind with significantly lower performances of 0.43 and 0.3, respectively. © 2017 Arribas-Bel et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.Ítem Seasonal analysis of the atmosphere during five years by using microwave radiometry over a mid-latitude site(Elsevier Ltd, 2019-01-01) Bedoya-Velásquez A.E.; Navas-Guzmán F.; de Arruda Moreira G.; Román R.; Cazorla A.; Ortiz-Amezcua P.; Benavent-Oltra J.A.; Alados-Arboledas L.; Olmo-Reyes F.J.; Foyo-Moreno I.; Montilla-Rosero E.; Hoyos C.D.; Guerrero-Rascado J.L.; Universidad EAFIT. Departamento de Ciencias Básicas; Óptica AplicadaThis work focuses on the analysis of the seasonal cycle of temperature and relative humidity (RH) profiles and integrated water vapor (IWV) obtained from microwave radiometer (MWR) measurements over the mid-latitude city of Granada, southern Spain. For completeness the study, the maximum atmospheric boundary layer height (ABLHmax) is also included. To this end, we have firstly characterized the HATPRO-RPG MWR errors using 55 co-located radiosondes (RS) by means of the mean-bias (bias¯) profile and the standard deviation (SDbias) profile classified under all-weather conditions and cloud-free conditions. This characterization pointed out that temperature from HATPRO-MWR presents a very low bias¯ respects RS mostly below 2.0 km agl, ranging from positive to negative values under all-weather conditions (from 1.7 to -0.4 K with SDbias up to 3.0 K). Under cloud-free conditions, the bias was very similar to that found under all-weather conditions (1.8 to -0.4 K) but with smaller SDbias (up to 1.1 K). The same behavior is also seen in this lower part (ground to 2.0 km agl) for RH. Under all-weather conditions, the mean RH bias ranged from 3.0 to -4.0% with SDbias between 10 and 16.3% while under cloud-free conditions the bias ranged from 2.0 to -0.4% with SDbias from 0.5 to 13.3%. Above 2.0 km agl, the SDbias error increases considerably up to 4 km agl (up to -20%), and then decreases slightly above 7.0 km agl (up to -5%). In addition, IWV values from MWR were also compared with the values obtained from the integration of RS profiles, showing a better linear fit under cloud-free conditions (R2 = 0.96) than under all-weather conditions (R2 = 0.82). The mean bias under cloud-free conditions was -0.80 kg/m2 while for all-weather conditions it was -1.25 kg/m2. Thus, the SDbiasfor all the statistics (temperature, RH and IWV) of the comparison between MWR and RS presented higher values for all-weather conditions than for cloud-free conditions ones. It points out that the presence of clouds is a key factor to take into account when MWR products are used. The second part of this work is devoted to a seasonal variability analysis over five years, leading us to characterize thermodynamically the troposphere over our site. This city atmosphere presents a clear seasonal cycle where temperature, ABLHmax and IWV increase from winter to summer and decrease in autumn, meanwhile RH decreases along the warmer seasons. This city presents cold winters (mean daily maximum temperature: 10.6 ± 1.1 °C) and dry/hot summers (mean daily maximum temperature of 28.8 ± 0.9 °C and mean daily maximum of surface RH up to 55.0 ± 6.0%) at surface (680 m asl). Moreover, considering temporal trends, our study pointed out that only temperature and RH showed a linear increase in winters with a mean-rate of (0.5 ± 0.1) °C/year and (3.4 ± 1.7) %/year, respectively, from ground to 2.0 km agl, meanwhile IWV presented a linear increase of 1.0 kg·m-2/year in winters, 0.78 kg·m-2/year in summers and a linear decrease in autumns of -0.75 kg·m-2/year. © 2018 Elsevier B.V.Ítem Si te gusta el café, tienes que ver esto(2020-12-01) Martinez Guerrero, Christian Alexander; Christian Alexander Martinez-Guerrero; Velásquez, David; Sánchez, Alejandro; Sarmiento, Sebastián; Toro, Mauricio; Maiza, Mikel; Sierra, Basilio; Estudios en Mantenimiento; GIDITICÍtem ¿Te gusta un buen café? Este hongo lo puede destruir(2020-12-01) Martinez Guerrero, Christian Alexander; Christian Alexander Martinez-Guerrero; Vasquez, David; Sanchez, Alejandro; Sarmiento, Sebastian; Toro, Mauricio; Maiza, Mikel; Sierra, Basilio; Vicerrectoría de Descubrimiento y CreaciónÍtem Using remote sensing to assess the relationship between crime and the urban layout(ELSEVIER SCI LTD, 2014-12-01) Patino, Jorge E.; Duque, Juan C.; Pardo-Pascual, Josep E.; Ruiz, Luis A.; Universidad EAFIT. Departamento de Economía y Finanzas; Research in Spatial Economics (RISE)The link between place and crime is at the base of social ecology theories of crime that focus in the relationship of the characteristics of geographical areas and crime rates. The broken windows theory states that visible cues of physical and social disorder in a neighborhood can lead to an increase in more serious crime. The crime prevention through environmental design (CPTED) planning approach seeks to deter criminal behavior by creating defensible spaces. Based on the premise that a settlement's appearance is a reflection of the society, we ask whether a neighborhood's design has a quantifiable imprint when seen from space using urban fabric descriptors computed from very high spatial-resolution imagery. We tested which land cover, structure and texture descriptors were significantly related to intra-urban homicide rates in Medellin, Colombia, while controlling for socioeconomic confounders. The percentage of impervious surfaces other than clay roofs, the fraction of clay roofs to impervious surfaces, two structure descriptors related to the homogeneity of the urban layout, and the uniformity texture descriptor were all statistically significant. Areas with higher homicide rates tended to have higher local variation and less general homogeneity; that is, the urban layouts were more crowded and cluttered, with small dwellings with different roofing materials located in close proximity to one another, and these regions often lacked other homogeneous surfaces such as open green spaces, wide roads, or large facilities. These results seem to be in agreement with the broken windows theory and CPTED in the sense that more heterogeneous and disordered urban layouts are associated with higher homicide rates. © 2014 Elsevier Ltd.