Álvaro Muriel Lima Machado

A Study on In Situ Calibration of an Off-The-Shelf Digital Camera Integrated to a Lidar System

Photogrammetric procedures using the integration of imagery and light detection and ranging (Lidar) datasets are becoming common in mapping companies nowadays. When photogrammetric and Lidar surveys are performed simultaneously while having the camera connected to the Lidar system, the images exterior orientation parameters computed from the Lidar system can be used to perform photogrammetric applications. To implement such approach, camera calibration procedures should be applied to obtain accurate interior orientation parameters. This paper shows the performed study and experimental results from an in situ self-calibration using simultaneously collected imagery and Lidar datasets. The experimental study is conducted to devise a methodology for the in situ self-calibration that uses a set of vertical control points, extracted from Lidar point cloud, as well as the three-dimensional coordinates of the camera exposure stations, computed from the Lidar trajectory. For the small photogrammetric block used in this study, the results from the performed experiments demonstrated that the in situ self-calibration is required to obtain horizontal accuracies of approximately one image pixel on the ground in bundle adjustment assisted by direct measured of the coordinates of the camera exposure stations.

Avaliação da varredura laser terrestre em diferentes distâncias da árvore para mensurar variáveis dendrométricas

O laser scanner terrestre e uma alternativa para a coleta de dados dendrometricos em campo sem a necessidade da derrubada da arvore. Assim, este estudo teve como objetivo avaliar a influencia da distância da varredura laser terrestre na determinacao das variaveis dendrometricas. Foram analisadas duas arvores com altura total de 17,14 m e 16,00 m. A varredura laser foi realizada a 5, 10, 15 e 20 m de distância da arvore. Os diâmetros foram obtidos em alturas parciais ate o topo da arvore. Os resultados obtidos com a varredura laser foram validados com as medidas obtidas tradicionalmente em campo com suta e trena. A melhor distância para a obtencao das variaveis dendrometricas foi de 15 m para ambas as arvores. Asxa0 distâncias menores que 10 m nao foram indicadas para a mensuracao das duas arvores analisadas. A determinacao da altura total foi coerente com as medidas parametricas. Foi possivel a determinacao de diâmetros ate 55% da altura total e, acima, a ramificacao da arvore resultou em uma super ou subestimativa. A melhor distância de varredura possivelmente esta atrelada a altura da arvore.

UM ALGORITMO PARA FILTRAGEM DO TRONCO EM NUVEM DE PONTOS LASER TERRESTRE DE ÁRVORES DE PINUS spp.

A varredura laser terrestre vem sendo testada e apontada como alternativa nao destrutiva para mensuracao de arvores. Variaveis dendrometricas podem ser obtidas a partir de tecnicas de modelagem tridimensional. Porem, e necessaria a filtragem dos dados para eliminar pontos que nao representam a superficie do tronco. Esse trabalho teve como objetivo apresentar uma proposta de algoritmo, baseada em metodos ja consolidados na literatura, para filtragem automatica da nuvem de pontos do tronco da arvore. O estudo foi conduzido a partir de doze arvoresxa0 inseridas em um povoamento inequiâneo de Pinus spp. Tres estacoes laser foram utilizadas para o recobrimento de cada arvore. O algoritmo proposto, denominado de Filtro Distância Maxima (Filtro Dmax), realiza a filtragem do tronco de forma iterativa e em secoes da nuvem de pontos. Para o melhor desempenho da automatizacao optou-se por um algoritmo parametrizado. O Filtro Dmax foi testado experimentalmente e os resultados apontaram que as superficies dos troncos foram efetivamente filtradas ate aproximadamente dois tercos da altura total das arvores estudadas. A alta densidade de aciculas impediu a representacao das superficies dos troncos no terco superior das arvores. A aplicacao do Filtro Dmax nao alterou as caracteristicas geometricas dos troncos proporcionando dados a modelagem tridimensional.

GeoGrid3D: a program to insert faults in 3D geological model for fluid simulation

This Paper show an application of geological and geophysical data integration into a flow simulation 3D model. A free program was used to do this flow simulation (Boast from United State Department of Energy). But to use this free simulator, a new program was build (GeoGrid3D). This program came with a special feature that can insert faults into the 3D model and also have other functions to get specific data codifications for Boast recognition. Outcrops from Permian sedimentary rocks and shallow seismic are used as data for this reservoir framework 3D model. Petrophysical properties were distributed using rules from interpretation of correlation done with facies and stratigraphic analysis of the outcrops. The main petrophysical model was built, and to test GeoGrid3D functions, other tree tectonic situations of faults comportament are also generated. The gas saturation, after flow simulation, confirms that changing faults permeabilities inside the reservoir will change gas production in rocks like this model.

Comparative analysis of the performance of metric-analog cameras, amateur-digital cameras, and LIDAR

The adaptation of digital technologies in today’s world is witnessing vast expansion where mapping tools are no exception. The increasing demand for rapid updates of spatial databases and the need for faster mapping end products are pushing towards using fast and cost-effective technologies. In traditional photogrammetric mapping, accurately calibrated analog metric cameras have been used to capture overlapping film photographs. Metric cameras are designed to provide extremely-high geometric image quality. They employ a low distortion lens system held in a fixed position relative to the film plane. The disadvantage of such systems is the high initial procurement cost and extra processing of photographs before measurements. The advent of low-cost off-the-shelf digital cameras encouraged both researchers and mapping companies to exploit such cameras in the mapping cycle. These cameras should be accurately calibrated and their accuracy and range of application should be determined. Besides photogrammetry, the evolving LIDAR (LIght Detection and Ranging) technology provides a new alternative for fast digital mapping. Based on a laser scanner and GPS/INS systems, a LIDAR system produces accurate point cloud measurements of surfaces and sometimes additional intensity images. Typical applications of LIDAR span mapping forestry floors, determination of power line sags, monitoring of coastal zones, city modeling, and construction surveys. The aim of this paper is to investigate the fit between the three mapping alternatives; metric-analog camera, low-cost/non-metric digital camera, and LIDAR. Two different types of cameras were used; Wild RC10 photogrammetric camera and Kodak 14n. Each camera was calibrated using a different calibration methodology and the number and arrangement of images taken were also different. As for the LIDAR dataset, an OPTECH ALTM 2050 laser scanner was used. Data from the laser range and reflected intensity were recorded. The comparative performance analysis is based on the quality of fit of the final alignment between the LIDAR and photogrammetric models through check-point analysis and derived orthophotos.