Submitted by Raquel Alves de Oliveira (raquel88@gmail.com) on 2017-12-07T10:06:49Z No. of bitstreams: 1 Oliveira_2017_TESE.pdf: 10400710 bytes, checksum: 4c4e6b235bd849c0d16074edea702847 (MD5) / Approved for entry into archive by ALESSANDRA KUBA OSHIRO null (alessandra@fct.unesp.br) on 2017-12-07T11:22:22Z (GMT) No. of bitstreams: 1 oliveira_ra_dr_prud.pdf: 10400710 bytes, checksum: 4c4e6b235bd849c0d16074edea702847 (MD5) / Made available in DSpace on 2017-12-07T11:22:23Z (GMT). No. of bitstreams: 1 oliveira_ra_dr_prud.pdf: 10400710 bytes, checksum: 4c4e6b235bd849c0d16074edea702847 (MD5) Previous issue date: 2017-06-29 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) / Recentemente, os sensores hiperespectrais miniaturizados entraram no mercado e alguns modelos adquirem bandas hiperespectrais com geometria de quadro, com a vantagem de serem também operados em veículos aéreos remotamente pilotados (VARP). As imagens deste tipo de câmara podem ser utilizadas para a geração de modelos digitais de superfície hiperespectral (MDSHs) de alta resolução, usando o VARP, sem a necessidade do registro de dados de diferentes sensores ou diferente datas de aquisição. MDSHs aumentam o conhecimento sobre os alvos, uma vez que permitem modelar a reflectância do alvo utilizando dados provenientes de diferentes direções. Neste trabalho, a câmara hiperespectral de quadro utilizada não adquire todas as bandas instantaneamente, causando um deslocamento entre as bandas devido ao movimento da plataforma. Os principais objetivos deste projeto foram estudar e desenvolver técnicas para a geração de MDSHs em áreas de florestas, investigando e avaliando as principais etapas para o processamento das imagens da câmara hiperespectral de quadro até a geração do MSDH. Considerando que a tecnologia da câmara baseia-se em filtros ajustáveis, o estudo avaliou: a auto-calibração da câmara, verificando o comportamento dos parâmetros de orientação interior em diferentes bandas espectrais; o corregistro das bandas através de transformações geométricas 2D; e a estimativa dos parâmetros de orientação exterior. Em relação à geração do MDS, uma abordagem baseada em correspondência de imagem no espaço do objeto foi desenvolvida, adaptando o método de busca em linha vertical (VLL) para a geração MDSH e foi nomeado como VLL hiperespectral (HVLL). Adicionalmente, o uso de imagens classificadas para a adaptação dos parâmetros de correspondência foi avaliado com o objetivo de melhorar o processo de correspondência para diferentes objetos (HVLLC). Posteriormente, foram utilizadas múltiplas bandas no processo de correspondência de imagens, dados como múltiplos ângulos de visada e informação espectral foram calculados simultaneamente ao processo de correspondência de imagens. A avaliação da qualidade foi realizada comparando-se os MDSs gerados com os produzidos por um software comercial e por dados Airborne Laser Scanning (ALS). Esta investigação demonstrou que a técnica proposta pode ser usada para a geração de modelos 3D integrados aos dados hiperespectrais multiangulares da câmara hiperespectral de quadro. A avaliação de todas as etapas demonstrou que esta tecnologia pode fornecer dados geométricos e espectrais precisos e os MDSHs resultantes possuem potencial para várias aplicações de sensoriamento remoto. / Recently, miniaturized hyperspectral sensors, operable from small Remotely Piloted Aerial Systems (RPAS), have entered the market and some of these sensors acquire hyperspectral bands in frame geometry. Images of the lightweight hyperspectral 2D frame camera can be used to generate high-resolution hyperspectral digital surface models (HDSMs), without the registration of data from different sensors or different dates of acquisition. HSDMs increase the knowledge about the targets since it allows modeling the target reflectance using data coming from different directions. In this study, the hyperspectral 2D frame camera used does not acquire all bands instantaneously, causing band misalignment due to the platform motion. The main aims of this project were to study and develop techniques for the generation of HDSMs in forest areas, studying and assessing the main steps to process the hyperspectral 2D frame camera images until the HDSM generation. Considering that the camera technology is based on tunable filters, the study have assessed the orientation and DSM generation steps: the self-calibrating bundle adjustment to verify the behaviour of the interior orientation parameters using different spectral bands; the co-registration of the bands using 2D geometric transformation; the exterior orientation parameter estimation. Regarding to the DSM generation, an approach based on object space image matching was developed, adapting the vertical line locus (VLL) method for HDSM generation, and was named as hyperspectral VLL (HVLL). Additionally, the use of image classification data was investigated in order to adapt the image matching parameters and improve the process of image matching for different objects (hyperspectral VLL classes - HVLLC). Further, multiple bands were used and the spectral and multiangular viewing geometry were computed simultaneously to the image matching method. Quality assessment was performed by comparing to DSMs generated to those produced by commercial software and also by Airborne Laser Scanning (ALS) data. This investigation demonstrated that the proposed technique can be used to generate integrated 3D information and multiangular hyperspectral data from hyperspectral 2D frame camera. The assessment of all steps showed that the hyperspectral 2D frame technology can provide accurate geometric and spectral data and the resulting HDSMs have potential for several remote sensing applications. / FAPESP: 2013/17787-3 / FAPESP: 2013/14444-0 / FAPESP: 2014/24844-6

<p>Unmanned aerial vehicles (UAVs) equipped with imaging systems and integrated global navigation satellite system/inertial navigation system (GNSS/INS) are used for a variety of applications. Disaster relief, infrastructure monitoring, precision agriculture, and ecological forestry growth monitoring are among some of the applications that utilize UAV imaging systems. For most applications, accurate 3D spatial information from the UAV imaging system is required. Deriving reliable 3D coordinates is conditioned on accurate geometric calibration. Geometric calibration entails both spatial and temporal calibration. Spatial calibration consists of obtaining accurate internal characteristics of the imaging sensor as well as estimating the mounting parameters between the imaging and the GNSS/INS units. Temporal calibration ensures that there is little to no time delay between the image timestamps and corresponding GNSS/INS position and orientation timestamps. Manual and automated spatial calibration have been successfully accomplished on a variety of platforms and sensors including UAVs equipped with frame and push-broom line cameras. However, manual and automated temporal calibration has not been demonstrated on both frame and line camera systems without the use of ground control points (GCPs). This research focuses on manual and automated spatial and temporal system calibration for UAVs equipped with GNSS/INS frame and line camera systems. For frame cameras, the research introduces two approaches (direct and indirect) to correct for time delay between GNSS/INS recorded event markers and actual time of image exposures. To ensure the best estimates of system parameters without the use of ground control points, an optimal flight configuration for system calibration while estimating time delay is rigorously derived. For line camera systems, this research presents the direct approach to estimate system calibration parameters including time delay during the bundle block adjustment. The optimal flight configuration is also rigorously derived for line camera systems and the bias impact analysis is concluded. This shows that the indirect approach is not a feasible solution for push-broom line cameras onboard UAVs due to the limited ability of line cameras to decouple system parameters and is confirmed with experimental results. Lastly, this research demonstrates that for frame and line camera systems, the direct approach can be fully-automated by incorporating structure from motion (SfM) based tie point features. Methods for feature detection and matching for frame and line camera systems are presented. This research also presents the necessary changes in the bundle adjustment with self-calibration to successfully incorporate a large amount of automatically-derived tie points. For frame cameras, the results show that the direct and indirect approach is capable of estimating and correcting this time delay. When a time delay exists and the direct or indirect approach is applied, horizontal accuracy of 1–3 times the ground sampling distance (GSD) can be achieved without the use of any ground control points (GCPs). For line camera systems, the direct results show that when a time delay exists and spatial and temporal calibration is performed, vertical and horizontal accuracy are approximately that of the ground sample distance (GSD) of the sensor. Furthermore, when a large artificial time delay is introduced for line camera systems, the direct approach still achieves accuracy less than the GSD of the system and performs 2.5-8 times better in the horizontal components and up to 18 times better in the vertical component than when temporal calibration is not performed. Lastly, the results show that automated tie points can be successfully extracted for frame and line camera systems and that those tie point features can be incorporated into a fully-automated bundle adjustment with self-calibration including time delay estimation. The results show that this fully-automated calibration accurately estimates system parameters and demonstrates absolute accuracy similar to that of manually-measured tie/checkpoints without the use of GCPs.</p>