Examinando por Autor "Forero Vargas, Manuel Guillermo"
Mostrando1 - 12 de 12
Resultados por página
Opciones de clasificación
- PublicaciónSólo datosA Note on the Phase Congruence Method in Image Analysis(Lecture Notes in Computer Science, 2019-03-03) Forero Vargas, Manuel Guillermo; Jacanamejoy, Carlos A.Phase congruence technique developed by Kovesi allows the detection of edges in images by analyzing the phases of their frequency components. A limitation of this technique is that it does not allow the detection of closely spaced edges that have different intensities. However, this situation occurs frequently in images, which therefore limits the use of this method. This study aims to propose a method that can overcome this limitation. Unlike the original technique, the proposed study uses a high degree of overlap between different frequency components to allow the detection of contiguous edges of low intensity. To avoid the problems that arise from high overlap, we modify the sensitivity of the phase congruence, allowing us to detect weak edges while discarding the noise associated with the proposed changes. We present our results and compare them with the results obtained using the existing technique.
- PublicaciónAcceso abiertoAnálisis comparativo de las técnicas SURF y ORB para la detección de puntos de interés en fotografías aéreas(Universidad de Ibagué, 2019) Godoy Olivera, Yesmar Andrés; Ducuara Oyuela, Álvaro Isledier; Forero Vargas, Manuel GuillermoThis project is part of a research developed, within the image processing and pattern recognition seedbed Lún the University of Ibagué, with the goal of evaluating detection and description of interest points techniques to register visible and near infrared aerial photographs. These pictures were acquired by the International Center for Tropical Agriculture "CIAT". The evaluated techniques are two recently published methods: "Speeded Up Robust Features", known as SURF, and "Oriented FAST and Rotated BRIEF", known by its acronym ORB, with the purpose of finding interest points employed as reference to merge them. Both methods were studied and a comparative analysis was carried out to determine which one is the most appropriate to merge this kind of images. In addition, Gabor filters were employed to find texture descriptors in the images, finding that entropy provides the best result.
- PublicaciónSólo datosAnalytical Comparison of Histogram Distance Measures(Lecture Notes in Computer Science, 2019-03-03) Forero Vargas, Manuel Guillermo; Arias-Rubio, Carlos; Tatiana González, BrigeteThis paper presents a comparative study of different distance measures used to compare histograms in applications such as pattern recognition, feature selection, image sorting, grouping, identification, indexing, and retrieval. The focus of the study is on how distance measures are affected by variations across images. Different distances between histograms were investigated and tested to compare their performance in retrieving gray scale and color images. A wide range of review papers on calculating distances between histograms was examined. One comparative study was found where histogram bins having zero value were discarded in the calculus of certain distances. We show that this is an inappropriate approach; our tests revealed that zero-value bins should be included to avoid erroneous calculations and achieve a performance advantage over other distance measures.
- PublicaciónAcceso abiertoClasificación automática de Heliconias a partir de imágenes RGB(Universidad de Ibagué., 2020) González Santos, Christian Saúl; Forero Vargas, Manuel Guillermo; Beltran Reyes, Carlos EduardoColombia is the country with the largest number of plant species in the world. Within it, the heliconias play an important ecological role within the ecosystems, since they are frequent components of the interior and limits of the forests, as well as of open environments such as pastures, roadsides and riverbanks. In some ecosystems they act as pioneers in the process of natural regeneration of vegetation and restoration of degraded soil. In addition, they maintain important co-evolutionary relationships with other animal and plant species, becoming an important element within the complex framework of life in the tropics. The classification of plant species is crucial for the protection and conservation of biodiversity. Manual classification is time-consuming, costly and requires experts who are often limited in their availability. To address these problems, three methods of classification of SVM (Support Vector Machine), ANN (Neural Networks), KNN (Nearest Neighbors) images with Euclidean distance and intersection were used in this work, which gave good results in the classification of four species of heliconias found at the University of Ibagué. The data used for training, testing and validation of the methods were RGB images taken in the natural habitat of the heliconias, in order to have information from their germination to their optimal cutting time. The images were pre-processed, making an adjustment of white balance, contrast and color temperature. To separate the heliconias from the background, a graphical segmentation technique using GPS was used. The descriptors were obtained using the technique known as BoW (Bag of Words), finding that the number of visual words most suitable for classification was between 20 and 30. The method with which the best results were obtained was the KNN; using the three closest neighbors, with an accuracy of 97%
- PublicaciónSólo datosColor Classification Methods for Perennial Weed Detection in Cereal Crops(Lecture Notes in Computer Science, 2019-03-03) Forero Vargas, Manuel Guillermo; Herrera-Rivera, Sergio; Ávila-Navarro, Julián; Franco, Camilo Andres; Rasmussen, Jesper; Nielsen, JonCirsium arvense is an invasive plant normally found in cold climates that affects cereal crops. Therefore, its detection is important to improve crop production. A previous study based on the analysis of aerial photographs focused on its detection using deep learning techniques and established methods based on image processing. This study introduces an image processing technique that generates even better results than those found with machine learning algorithms; this is reflected in aspects such as the accuracy and speed of the detection of the weeds in the cereal crops. The proposed method is based on the detection of the extreme green color characteristic of this plant with respect to the crops. To evaluate the technique, it was compared to six popular machine learning methods using images taken from two different heights: 10 and 50 m. The accuracy obtained with the machine learning techniques was 97.07% at best with execution times of more than 2 min with 200 × 200-pixel subimages, while the accuracy of the proposed image processing method was 98.23% and its execution time was less than 3 s.
- PublicaciónAcceso abiertoEstimación de fenotipos de plántulas de maíz a partir de nubes de puntos adquiridas en ambiente controlado(Universidad de Ibagué, 2022) Betancourt Lozano, Juan Jose; Forero Vargas, Manuel Guillermo; Murcia Moreno, Harold FabianLa fenotificación o estimación de rasgos morfológicos en plántulas, mediante técnicas se miautomáticas, es cada vez más relevante en los estudios de valoración de genotipos a nivel de órganos, consiguiendo tener un análisis más preciso de las características de ca da planta en ambientes controlados, permitiendo de esta manera encontrar las mejores variedades. Así, en este trabajo se presenta una técnica semiautomática para la segmen tación de órganos de plántulas de maíz y la estimación de características fenotípicas a nivel poblacional, individual y de órganos. Para ello, se utilizó una plataforma LiDAR construida previamente por otro miembro del proyecto y la base de datos obtenida en dicho trabajo como punto de partida. Mediante pruebas experimentales, se estableció un protocolo de adquisición para mejorar la calidad de los datos obtenidos por la plataforma, el cual se utilizó para aumentar la base de datos existente. Luego, se realizó un estudio comparativo de filtros empleados para la eliminación de ruido en nubes de puntos con el fin de mejorar la calidad de la base de datos. Posteriormente, empleando técnicas de aprendizaje de maquinas clásico y procesamiento de imágenes, se realizó la segmentación del tallo de la planta y la identificación de cada hoja. Los resultados obtenidos arrojaron una exactitud del 85.71 % en la segmentación del tallo, 85.50 % en el conteo de hojas y 70.11 % en la identificación de hojas. Por último, se obtuvieron algunos fenotipos del maíz que se encuentran en la literatura y se analizaron a nivel poblacional (altura promedio y cantidad de hojas promedio), individual (altura, área foliar, ángulo de inclinación y volumen) y de órganos (largo, ancho, ángulo de inclinación y volumen).
- PublicaciónAcceso abiertoEstudio de movimiento de objetos en imágenes basados en el método de flujo óptico de Horn y Schunck(Universidad de Ibagué., 2019) Briñez Varon, Cristian Javier; Forero Vargas, Manuel GuillermoOne of the most interesting topics in the area of computer vision is the analysis of the movement of objects in a video sequence, a technique commonly known as optical flow. Although several methods have been proposed to solve the problem of optical flow, there is no optimal solution since it is a poorly planned problem, because the video images are only a two-dimensional projection of a 3D world. In this work a study of the original technique of Horn and Schunck is made that gave origin to the investigations in this area, and the use of multi-scale representations that have been used recently to improve the characteristics of the original technique. The multi-scale method is based on reducing the size of the image successively, thus a multi-level pyramid is constructed, starting from the original size or scale of each input image, in order to detect the smallest displacement, since that the original technique fails to detect movements larger than the size of a pixel. The algorithm used by this technique detects first the largest movements and the result is used as the starting point to detect the shortest movements, handling a pyramidal structure based on the previous procedure to calculate the movement at all the scales.
- PublicaciónAcceso abiertoEvaluacion de filtros no lineales basados en los métodos Rof y Anisotropo para la eliminación de ruido en imágenes de microscopia confocal(Universidad de Ibagué., 2019) Pena Ambrocio, Reynel Duvan; Forero Vargas, Manuel GuillermoThe Parhyale hawaiensis is a useful model to understand the characteristic events of tissue regeneration and also has great experimental advantages, since with a confocal microscopy equipment and the help of programs for the analysis of images it is possible to quantify cell growth and the appearance of new cells. However, this process is highly routine, inefficient in terms of time and resources, since the processing of images must be done manually in some relevant software for image analysis. Therefore, new methods of image processing are required to automate these processes or make them more efficient, in order to improve research, allowing impartial, reliable and reproducible results. The first step for the development and identification of cells is to eliminate the noise, without affecting its edges, because the signal produced is very weak and the images present a large amount of Poisson noise. This step is necessary in the development of a technique for monitoring cells in microscopy images, required to advance research and knowledge of the mechanisms of regeneration in animals. Due to their great importance, a significant number of non-linear filters have been developed, some of which have not been sufficiently explored for use in microscopy. This research focuses on filters based on anisotropic and ROF. In addition, a comparative analysis of the anisotropic diffusion and ROF filters with the median, average, guided, propagated and propagated filters are presented to determine which of this have the best results in microscopy and synthetic images, developed to determine the quality of the filters.
- PublicaciónSólo datosGenerating Random Variates via Kernel Density Estimation and Radial Basis Function Based Neural Networks(Lecture Notes in Computer Science, 2019-03-03) Forero Vargas, Manuel Guillermo; Herrera-Rivera, Sergio; Candia-García, CristianWhen modeling phenomena that cannot be studied by deterministic analytical approaches, one of the main tasks is to generate random variates. The widely-used techniques, such as the inverse transformation, convolution, and rejection-acceptance methods, involve a significant amount of statistical work and do not provide satisfactory results when the data do not conform to the known probability density functions. This study aims to propose an alternative nonparametric method for generating random variables that combines kernel density estimation (KDE), and radial basis function based neural networks (RBFBNNs). We evaluate the method’s performance using Poisson, triangular, and exponential probability density distributions and assessed its utility for unknown distributions. The results show that the model’s effectiveness depends substantially on selecting an appropriate bandwidth value for KDE and a certain minimum number of data points to train the algorithm. the proposed method enabled us to achieve an R2 value between 0.91 and 0.99 for analyzed distributions.
- PublicaciónAcceso abiertoMétodo automático para la detección de quistes en el cerebro de cerdos a partir de imágenes de resonancia magnética(Universidad de Ibagué., 2019) Hatty Ramírez, Luisa María; Barragán Aya, Yesid Armando; Forero Vargas, Manuel GuillermoThe digital processing of images applied to the biomedical area has become an indispensable tool for better medical information extracting, assisting reliable diagnosis. Neurocysticercosis is an infectious disease known to cause the presence of cysts in the brain. The most frequent symptoms are convulsions and headache. The study of drugs to combat this disease is carried out on animal models such as pigs, given that for ethical reasons it is not originally possible to do it on human being. Normally, the study is done on microscopic images of different cuts of the brain of the animal. Recently, the use of magnetic resonance imaging has been proposed for study, but the analysis is generally complex due to unwanted pig movements, which does not allow high-resolution acquisitions. Currently, there is no method that allows a quick and efficient analysis from the magnetic resonance images of the pig's head in order to obtain valuable information to prove the effectiveness of medicines in the treatment of this disease; thus the analyses are performed manually, which takes a great amount of effort and time. Hence the importance of designing a method based on image processing techniques to identify the amount of cysts presents in the brain and its volume to evaluate the effectiveness of the treatment. Thus, this work describes the method developed in Java programming language applied to the free image processing software ImageJ to obtain the number and volume of cysts.
- PublicaciónSólo datosSegmentation of Meristem Cells by an Automated Optimization Algorithm(Applied Sciences, 2020-11-28) Rojas, Oswaldo; Forero Vargas, Manuel Guillermo; Menéndez, José Manuel; Jones, Angharad; Dewitte, Walter; Murray, JamesMeristem cells are irregularly shaped and appear in confocal images as dark areas surrounded by bright ones. Images are characterized by regions of very low contrast and absolute loss of edges deeper into the meristem. Edges are blurred, discontinuous, sometimes indistinguishable, and the intensity level inside the cells is similar to the background of the image. Recently, a technique called Parametric Segmentation Tuning was introduced for the optimization of segmentation parameters in diatom images. This paper presents a PST-tuned automatic segmentation method of meristem cells in microscopy images based on mathematical morphology. The optimal parameters of the algorithm are found by means of an iterative process that compares the segmented images obtained by successive variations of the parameters. Then, an optimization function is used to determine which pair of successive images allows for the best segmentation. The technique was validated by comparing its results with those obtained by a level set algorithm and a balloon segmentation technique. The outcomes show that our methodology offers better results than two free available state-of-the-art alternatives, being superior in all cases studied, losing 9.09% of the cells in the worst situation, against 75.81 and 25.45 obtained in the level set and the balloon segmentation techniques, respectively. The optimization method can be employed to tune the parameters of other meristem segmentation methods.
- PublicaciónAcceso abiertoSeparación y conteo de células musculares en procesamiento de imágenes(Universidad de Ibagué., 2019) Urrego Gamboa, Diego Alejandro; Español Díaz, Jorge Danilo; Forero Vargas, Manuel GuillermoIn medicine and pharmacology it is common to use animal models for the development and evaluation of new medications, since, by ethics, it is not possible to cause harm or experiment with humans. However, little has been done in biomedical engineering to facilitate the work of scientists for studies on these models. In particular, a difficulty that arises in the analysis of muscle tissue is the identification and counting of cells in microscopy images of lower rat limbs. This task is usually done manually, which makes it tedious, subjective, inaccurate, not reproducible and expensive in time. For this reason, in this work a semi-automatic technique was developed based on image processing techniques that allows to accelerate the cell counting process, significantly reducing the analysis time and increasing the accuracy. An editing tool was included to correct the results obtained. For the study, images provided by the Stem Cell Laboratory of the Department of Physiology and Pathological Transplantation of the University of Milan, Italy were used. The method implemented as a tool of the ImageJ program was evaluated and compared with Ground Truth images, obtaining a reduction in time of 83%.