Fernando Marson | Unisinos - Academia.edu (original) (raw)

Papers by Fernando Marson

We investigate here the potential of the spectroscopy in the identification and quantification of... more We investigate here the potential of the spectroscopy in the identification and quantification of mixtures of kaolinite and goethite from the Continuum Removal (CR) of the spectra in the short-wave infrared. For this purpose, spectral measurements of the kaolinite, goethite, and controlled mixtures of these minerals were systematically performed. The continuum of the results were removed, the depth of the kaolinite diagnostic absorption was calculated and compared with a spectral library. It was possible to identify the kaolinite with high determination coefficient (R2>0.95) when its proportion reaches at least 60% in the mixture. For quantification purposes, it was possible to quantify kaolinite using the diagnostic absorption feature depth in the CR with a coefficient of determination of 0.99.

In this study we present a mobile application for geoscience. It refers to a digital field book f... more In this study we present a mobile application for geoscience. It refers to a digital field book for automating data collection and outcrop/core description, and optimizing the final data processing. Sensors were developed for semi-automatic data collection, real time calculations, measurements of dip angles and dip directions, geographic location, among others. Field tests were performed comparing the traditional method with the proposed digital method. The preliminary results show a good acceptance of the mobile application by geoscientists and an improvement in the time required to perform data collection. Field tests are still going on and the complete results will be used to improve the development of this important tool in geoscience field work.

2018 IEEE Conference on Virtual Reality and 3D User Interfaces (VR)

Laser scanning and photogrammetry techniques have been broadly adopted by Oil&Gas industry fo... more Laser scanning and photogrammetry techniques have been broadly adopted by Oil&Gas industry for modeling petroleum reservoir analogues. Beyond the benefits of digital data itself, computer systems employed by geoscientists for interpretation and modeling tasks provide high quality rendering, point clouds surface meshes and photo-realistic textured models. But these systems, commonly, have used 2-D display, the 3-D models and information are projected on the screen, providing a limited visualization and restrictive toolset for interpretation. This work proposes to break this paradigm by developing a fully immersive system capable to virtually teleport the geoscientists to the fieldwork and provide a complete toolset for the outcrop's interpretation. Besides, the system has been evaluated and validated by geologists with different skills and it has emerged as an useful and attractive toolset for Oil&Gas industry.

2017 IEEE International Geoscience and Remote Sensing Symposium (IGARSS), 2017

IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium, 2018

This study aimed the analysis of the positional and geometric accuracy of objects in orthomosaics... more This study aimed the analysis of the positional and geometric accuracy of objects in orthomosaics obtained through different unmanned aerial vehicle (UAV) data processing software covering an area located within Universidade do Vale do Rio dos Sinos – UNISINOS in São Leopoldo, RS. A total of nine ground control points (GCP) and twenty checkpoints were surveyed in order register and classify the processed orthomosaics according to the cartographic accuracy standard – Padrão de Exatidão Cartográfica (PEC). Four software was employed to process the UAV data: Pix4D mapper, Agisoft PhotoScan, Menci APS and Bentley Context Capture. The results obtained from each software were compared and identified the smallest distortions when processing with and without ground control points. The flight was executed at a height of 90m with 60% sidelap and 80% overlap using an ST800 UAV equipped with a Sony NEX-7 small format non-metric camera with 24 megapixels resolution. The software GeoPEC was used to classify the orthomosaics according to PEC. For data processed with ground control points all orthomosaics were classified “Class A” in 1/500 scale, however, only Menci APS did not present a trend line via t-student test. On the other hand, Menci APS presented the worst results without the ground control points. In processing with GCP, all orthomosaics obtained optimum results with an approximated error of 2,5 m2, about 0,03% of the area.

2018 IEEE Conference on Virtual Reality and 3D User Interfaces (VR), 2018

The main goal of this paper was to evaluate the use of a low cost immersive driving simulator to ... more The main goal of this paper was to evaluate the use of a low cost immersive driving simulator to improve the teaching learning process of the Transport Infrastructure undergraduate course. The driving simulator that was developed in a virtual reality environment to assist both the teaching of engineering and the research on road safety. An experiment was conducted in Transport Infrastructure 1 course for Civil Engineering students in a Brazilian university. The students developed a geometric design of a road that was posteriorly modeled in 3D and provided in simulator. Students piloted a vehicle in the immersive simulator in the same road that they designed. Subsequently the usability of the system was assessed by the SUS metric (System Usability Scale). We performed an evaluation with 52 users and the SUS metric that we found was of 73% assuring a degree of usability above average and demonstrating that the immersive system is good to be used as a complementary tool in the learning of transport infrastructure.

2017 IEEE International Geoscience and Remote Sensing Symposium (IGARSS), 2017

In this study, we propose a method to calibrate the laser pulse return intensity of a Terrestrial... more In this study, we propose a method to calibrate the laser pulse return intensity of a Terrestrial Laser Scanner (TLS) based on Artificial Neural Networks. The laser pulse return intensity has an important rule on rocks types' classification when using Digital Outcrops Models (DOM) and has been the focus of much research by the geological community as it helps the geological interpretation in outcrops. In our experiment, we used a TLS Ilris 3D model with a wavelength of 1,535 nm. Our method has shown good efficiency for the calibration of the laser pulse return intensity, demonstrating a strong applicability for classification studies of rock types on Digital Outcrops Models.

2017 IEEE International Geoscience and Remote Sensing Symposium (IGARSS), 2017

Airborne and terrestrial laser scanning techniques have been largely used for the reconstruction ... more Airborne and terrestrial laser scanning techniques have been largely used for the reconstruction of high-resolution 3-D topography in the field of geosciences. In recent years, laser scanning has been also exploited on rock properties, biomass classification and carbon storage estimation. However, when laser spot collides partially against the target or even against undesirable background objects, part of emitted beam is lost and does not return to the laser station. So, it can introduce fewer discontinuities or even artifacts in the point cloud borders, comprising the results. Assuming an interest in minimizing this border effect, we have proposed a computational postprocessing algorithm which identifies anomalies and discrpancies and minimize it by recovering the expected intensity of returned laser pulse. The proposed technique operates on the basis of the collected point cloud intensity of return pulse, laser scanner's position and signals divergence, without requiring any kind of previous setup or additional accessory to the laser scanner.

This paper describes the development of a procedural pseudoinfinite deformable terrain system. Th... more This paper describes the development of a procedural pseudoinfinite deformable terrain system. The proposed method intends to deal with the high performance required by games. In this work the terrains are synthesised using Bezier curves and are loaded and saved automatically as needed, that is according to the camera position. The Heightmap’s generating process is, at some moments of a game, an expensive task. Considering such aspect we perform the Heightmap’s synthesis in the background of the application, restricting it to a specific FPS target. Moreover, we developed a deformation model which allows alterations in the HeightMap and Texture. We have used texture arrays at the GPU level to prevent sending unnecessary data. The performance of the system is measured both in the average FPS during the deformation of the terrain and the time taken for the synthesis

The use of procedural generation techniques has emerged as a growing trend in the field of digita... more The use of procedural generation techniques has emerged as a growing trend in the field of digital games due to the numerous applicability in content generation, as this has been shown as the main bottleneck in high-quality commercial titles production process, affecting mainly small developers with limited budget. Faced with this problem, this paper presents a general purpose procedural content generation framework for digital games, called gameBITS Framework, relying on a modular and hierarchical architecture for use in different environments and purposes. The objective of this work is to provide a simple, robust and extensible solution for small developers of digital games in low and medium complexity applications, and through this ensure cost reduction and production time. For this purpose, a review of studies in the field of procedural generation is presented to clarify the distinction between the various types of content and techniques applicable in this context as well as oth...

2017 IEEE International Geoscience and Remote Sensing Symposium (IGARSS), 2017

Computer vision applications have become popular in different areas of knowledge in recent years.... more Computer vision applications have become popular in different areas of knowledge in recent years. Augmented reality is an area that depends heavily on advances in computer vision techniques, and is necessary for the development of numerous applications for entertainment, education and business. One of the main challenges when working with computer vision is the need for controlled environments to enable image capture of acceptable quality. In this article, we present a model to recognize hand-drawn labyrinth designs from images captured with a webcam. We perform a luminosity calibration applying filters to detect the labyrinth’s edges. With this information, we then perform a sequence of steps to connect each pair of corners forming a line, resulting in points which can be used to create a 3D model through an extrusion process.

2017 IEEE International Geoscience and Remote Sensing Symposium (IGARSS), 2017

The use of LiDAR and multiples digital images jointly with 3-D reconstruction techniques for crea... more The use of LiDAR and multiples digital images jointly with 3-D reconstruction techniques for creating 3-D models of natural outcrops and surfaces studies have increased dramatically in the last few years. These techniques have provided an enormous amount of data for interpretation by geoscientists. However, these researchers have no available software capable of offering a user experience comparable to the fieldwork. The majority of solutions have considered desktop systems, which presents inherent limitations due to the 2-D characteristics of displays and loss of immersion into the 3-D model, or up until expensive and complex stereoscopic based approaches to improve the 3-D user experience do not offer well suitable solutions. To address these limitations, this paper presents a low-cost completely disruptive solution for processing, visualizing, sharing and directly handling Digital Outcrop Models with the support of a full interpretation toolset, the MOSIS System. The proposed system provides a fully immersive computational environment, capable of teleporting virtually geoscientists to the fieldwork, giving an awareness of being there physically with an extensible toolset for the DOM's interpretation. Besides, desktop, web and mobile versions of MOSIS have been under development and fulfill the lack of tools for digital outcrop modeling.

IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium, 2018

Computer vision systems allow digital reconstruction of targets by capturing information through ... more Computer vision systems allow digital reconstruction of targets by capturing information through remote sensors such as video cameras and scanners. In this context, the objective of this work was to evaluate the capacity and quality of three-dimensional reconstruction of static targets using the ZED stereoscopic camera. For this goal, we took images of several environments and objects with different surfaces, textures, lighting, distances and acquisition speeds. The results were compared with high-density and high precision point clouds obtained from the targets using a Leica Viva TS15 total station. The data were processed in the CloudCompare software to calculate the displacement between the models generated by the camera and the total station. Under certain circumstances, this technology is able to reconstruct three-dimensional objects and environments with an error of a few centimeters.

IEEE Geoscience and Remote Sensing Magazine, 2018

2015 14th Brazilian Symposium on Computer Games and Digital Entertainment (SBGames), 2015

This article presents a computational model for procedural generation of customized façade and in... more This article presents a computational model for procedural generation of customized façade and interior styles of buildings for use in three-dimensional virtual environments of games and simulations. The model makes use of two types of input information: geometric and semantic. The geometric information is related to the two-dimensional floor plan of a building, with its parts and dimensions, as well as positions for doors and windows. The semantic information enables the creation of architectural styles, enabling variations for materials and textures for the façade and inner parts, as well as for the shapes and dimensions of doors and windows. Changing one or more input parameters modifies the final appearance of the result. The proposed computational model can be used to generate large virtual environments, as it allows mixing different floor plans and architectural styles to achieve visual diversity. The main characteristics of the work are the realtime procedural generation of 3D buildings, the customization of façades and building interiors, as well as the use of semantic to assign meaning to the different elements of the house.

Anais dos Workshops do V Congresso Brasileiro de Informática na Educação (CBIE 2016), 2016

2018 IEEE Conference on Virtual Reality and 3D User Interfaces (VR)

2017 IEEE International Geoscience and Remote Sensing Symposium (IGARSS), 2017

IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium, 2018

2018 IEEE Conference on Virtual Reality and 3D User Interfaces (VR), 2018

2017 IEEE International Geoscience and Remote Sensing Symposium (IGARSS), 2017

IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium, 2018

IEEE Geoscience and Remote Sensing Magazine, 2018

2015 14th Brazilian Symposium on Computer Games and Digital Entertainment (SBGames), 2015

Anais dos Workshops do V Congresso Brasileiro de Informática na Educação (CBIE 2016), 2016