付記する学位プログラム名: 霊長類学・ワイルドライフサイエンス・リーディング大学院 / 京都大学 / 新制・課程博士 / 博士(理学) / 甲第24469号 / 理博第4968号 / 新制||理||1709(附属図書館) / 京都大学大学院理学研究科生物科学専攻 / (主査)教授 平田 聡, 教授 伊谷 原一, 教授 村山 美穂 / 学位規則第4条第1項該当 / Doctor of Science / Kyoto University / DGAM

Submitted by Cássia Santos (cassia.bcufg@gmail.com) on 2015-11-26T10:24:30Z No. of bitstreams: 2 Dissertação - Leomar Rufino Alves Júnior - 2015.pdf: 5358486 bytes, checksum: 65b930c6a83152c4867b2f00eb5865b2 (MD5) license_rdf: 23148 bytes, checksum: 9da0b6dfac957114c6a7714714b86306 (MD5) / Approved for entry into archive by Luciana Ferreira (lucgeral@gmail.com) on 2015-11-27T07:27:21Z (GMT) No. of bitstreams: 2 Dissertação - Leomar Rufino Alves Júnior - 2015.pdf: 5358486 bytes, checksum: 65b930c6a83152c4867b2f00eb5865b2 (MD5) license_rdf: 23148 bytes, checksum: 9da0b6dfac957114c6a7714714b86306 (MD5) / Made available in DSpace on 2015-11-27T07:27:21Z (GMT). No. of bitstreams: 2 Dissertação - Leomar Rufino Alves Júnior - 2015.pdf: 5358486 bytes, checksum: 65b930c6a83152c4867b2f00eb5865b2 (MD5) license_rdf: 23148 bytes, checksum: 9da0b6dfac957114c6a7714714b86306 (MD5) Previous issue date: 2015-03-13 / Fundação de Amparo à Pesquisa do Estado de Goiás - FAPEG / This research aimed to verify the precision and accuracy of orthomosaics and Digital Surface Model (DSM) generated automatically by aerial photography taken with an Unmanned Aerial Vehicle (UAV) in urban and rural areas at the cities of Goiânia, Goiás and Edéia, all located in the state of Goiás. The survey also verified for the influence of the scale, depending on the flight height, the influence of the sunlight, and the phenological analysis obtained in agricultural areas with sugarcane and corn (vegetative cycle) crops. Flight plans were drawn up in E-mo-tion software provided by Sensefly - Swiss company manufacturer of UAV Swinglet CAM used in this work. The camera on board UAV was the Canon IXUS 220 HS, with a spatial resolution of 12.1 megapixel, CMOS sensor equipped with type 1 / 2.3" (4000 x 3000 pixel), pixel pitch of 1.54 m, and focal distance equivalent of 35 mm. To check the precision and accuracy of orthomosaics in urban areas, the flights were uniformly distributed in the study area with three-dimensional coordinates pre-marked targets read in the orthomosaic itself, and compared with the coordinates obtained by RTK and static fast positioning methods geodetic survey, using a pair of GNSS signal receiver. Evaluation of Cartographic Accuracy Standards (PEC, defined by the Brazilian decree, no. 89817 of June 20, 1984) was performed by discrepancies between these coordinates. The bias was analyzed by student's t test and the accuracy with the chi-square probability. We have found that in those orthomosaics performed over the urban area in Goiânia city, some buildings were not properly processed in terms of the conical to orthogonal projection; this product was classified as PEC Class A to the 1/250 scale. In Goiás city, the generated orthomosaics without support points were classified as PEC Class A to the 1/2,500 scale, while the orthomosaic generated with eight supporting points was classified as PEC Class A to the 1/1,125 scale. The orthomosaic and MDS generated without ground supporting points presented planialtimetric trends. The mean difference calculated in the orthomosaic with ground supporting was 54 times lower on the E axis, and 111 times lower in N axis, and 10 times lower in the Z axis relative to the average of the discrepancies in orthomosaic without ground control points. The MDS generated with and without supporting showed vertical displacement trend. Thus it was evident the need for supporting points for making orthomosaics and MDS obtained with UAV. In flights performed in crop areas, it was observed that for a better estimate of MPRI vegetation index, the photographs needs to with larger scales (GSD of 5 cm), even with a higher software difficulty in finding homologous points. It did not occur when we used pictures with smaller scales (GSD 20 cm, or larger) for the generation of orthomosaics and MDS. / Esta pesquisa teve por objetivo verificar a precisão e acurácia dos ortomosaicos e Modelo Digital de Superfície (MDS) gerados automaticamente por programa de aerofotogrametria, utilizando fotografias aéreas tomadas com um Veículo Aéreo Não Tripulado (VANT) em áreas urbanas e rurais das cidades de Goiânia, Goiás e Edéia, todas no estado de Goiás. A pesquisa também verificou a influência da escala, em função da altura de voo, a influência da iluminação solar e a análise fenológica em ortomosaicos obtidos em áreas agrícolas com cultivo de cana-de-açúcar e milho (clico vegetativo). Os planos de voo foram elaborados no programa E-mo-tion, fornecido pela Sensefly – empresa suíça fabricante do VANT Swinglet CAM utilizado nesse trabalho. A câmara instalada no VANT foi a Canon IXUS 220 HS, com resolução espacial de 12,1 megapixel, equipada com sensor tipo CMOS 1/2,3” (4000 x 3000 pixel), pixel pitch de 1,54 m, e distância focal equivalente de 35 mm. Para verificar a precisão e acurácia dos ortomosaicos, nas áreas urbanas, foram uniformemente distribuídos nas áreas de estudo alvos pré-sinalizados, com coordenadas tridimensionais lidas nos ortomosaicos e comparadas com as coordenadas obtidas por levantamento geodésico nos métodos de posicionamento RTK e estático rápido, utilizando-se um par de receptor de sinais GNSS. A avaliação do Padrão de Exatidão Cartográfica (PEC) foi realizada pelas discrepâncias entre essas coordenadas. A tendenciosidade foi analisada pelo teste t de Student e a precisão pela probabilidade do qui-quadrado, considerando o ortomosaico Classe A conforme PEC estabelecido no Decreto nº 89.817 de 20.06.1984. Verificou-se que nos ortomosaicos oriundos do aerolevantamento realizado na área urbana (Goiânia), algumas edificações não foram devidamente transformadas da projeção cônica para a ortogonal. Os ortomosaicos gerados com mais de 8 pontos de controle foram classificados como Classe A para a escala de 1/250. Verificou-se, também, que o ortomosaico sem pontos de apoio, oriundo dos aerolevantamentos realizados na cidade de Goiás, foi classificado como Classe A na escala 1/2.500, enquanto o ortomosaico gerado com oito pontos de apoio foi classificado como Classe A na escala de 1/1.125, conforme parâmetros de precisão e exatidão estabelecidos por este mesmo Decreto Federal. O ortomosaico e MDS gerados sem pontos de apoio apresentaram tendência planialtimétrica. As médias das discrepâncias calculadas no ortomosaico com apoio foi 54 vezes menor no eixo E, 111 vezes menor no eixo N e 10 vezes menor no eixo Z, em relação à média das discrepâncias no ortomosaico sem pontos de apoio. O MDS gerado com e sem apoio apresentaram tendência de deslocamento vertical. Ficou evidente a necessidade de utilizar pontos de apoio para a confecção de ortomosaicos e MDS obtidos com VANT. Em relação aos voos realizados em áreas de cultivo observou-se que para uma melhor estimativa do MPRI observou-se que na geração de ortomosaicos a partir de fotografias com escalas maiores (GSD de 5 cm), o programa de processamento das fotografias teve dificuldade em encontrar os pontos homólogos necessários à geração dos ortomosaicos. Fato que não ocorreu quando se utilizou fotografias com escalas menores (GSD de 20 cm) para a geração dos ortomosaicos e MDS.

<p>Three prescribed burn sites and seven selective timber harvest sites were surveyed using a UAS equipped with a PPK-triggered RGB sensor to determine optimal image collection parameters surrounding each type of disturbance and land cover. The image coordinates were corrected with a third-party base station network (CORS) after the flight, and photogrammetrically processed to produce high-resolution georeferenced orthomosaics. This addressed the first objective of this study, which was to <i>establish effective data procurement methods from both before and after planned </i>disturbances. <br></p><p>Orthomosaic datasets surrounding both a prescribed burn and a selective timber harvest, were used to classify land covers through geographic image-based analysis (GEOBIA). The orthomosaic datasets were segmented into image objects, before classification with a machine-learning algorithm. Land covers for the prescribed prairie burn were 1) bare ground, 2) litter, 3) green vegetation, and 4) burned vegetation. Land covers for the selective timber harvest were 1) mature canopy, 2) understory vegetation, and 3) bare ground. 65 samples per class were collected for prairie burn datasets, and 80 samples per class were collected for timber harvest datasets to train the classifier. A supported vector machines (SVM) algorithm was used to produce four land cover classifications for each site surrounding their respective planned disturbance. Pixel counts for each class were multiplied by the ground sampled distance (GSD) to obtain area calculations for land covers. Accuracy assessments were conducted by projecting 250 equalized stratified random (ESR) reference points onto the georeferenced orthomosaic datasets to compare the classification to the imagery through visual interpretation. This addressed the second objective of this study, which was to <i>establish effective data classification methods from both before and after planned </i>disturbances.<br></p><p>Finally, a two-tailed t-Test was conducted with the overall accuracies for each disturbance type and land cover. Results showed no significant difference in the overall accuracy between land covers. This was done to address the third objective of this study which was to <i>determine if a significant difference exists between the classification accuracies between planned disturbance types</i>. Overall, effective data procurement and classification parameters were established for both <i>before </i>and <i>after </i>two common types of <i>planned </i>disturbances within the CHF region, with slightly better results for prescribed burns than for selective timber harvests.<br></p>