Macquarie Island Quickbird Image (15 March 2005) orthorectification

Quickbird image of Macquarie Island acquired on 15 March 2005. The image has been orthorectified, i.e. all geometric distortions caused by the earth curvature and relief displacement have been corrected. Twenty six ground control points spread out over the image were identified, based on survey control marks and corresponding pixels in the Panchromatic band of the Quickbird image. This was done in collaboration with Henk Brolsma, Australian Antarctic Data Centre.The image was transformed based on the Quickbird RPC coefficients (sensor platform parameters) and the ground control points. The 5m AIRSAR DEM acquired by NASA (metadata record - Macquarie Island AIRSAR DEM (Digital Elevation Model / Entry ID: macca_dem_gis ) was used to correct the image for topographic relief distortions. The DEM was resampled with the bilinear resampling algorithm; the image was resampled with the nearest neighbour algorithm to retain the original pixel values. The horizontal accuracy is estimated to be within 5m or 2 pixels, but in places this may be greater. Coordinate system: WGS84 datum UTM zone 57S projection.The orthorectification was done by Dr Arko Lucieer, Centre for Spatial Information Science (CenSIS), School of Geography and Environmental Studies, University of Tasmania using ENVI 4.3 and the new image was exported to a GeoTIFF and a JPEG2000 (lossless compression) image.Quickbird imagery consists of four multi-spectral bands: Blue, Green, Red, Near-Infrared at 2.4m resolution (pixel size). Quickbird also acquires a panchromatic image (grey scale) at 0.6m resolution.Two colour composites have been included:1. Visible bands (RGB = band 3, 2, 1) corresponding to the way the human visual system sees colours and very similar to an aerial photograph.2. False colour composite (RGB = band 4, 3, 2). This colour composite includes the Near-Infrared band (band 4) to highlight vegetation in red. Cholorphyll in vegetation causes a high reflectance of NIR wavelength energy which shows up as bright red in the image. Vegetation that has been grazed or dies back looses its red colour.Two image formats have been included:1. GeoTIFF (.tif and .tfw). This image is uncompressed and can be opened in ArcView 3.x or ArcMap without loading an extension.2. JPEG2000. This image format is highly compressed without loosing image quality. This image is exactly the same as the GeoTIFF but has a smaller file size. JPEG2000 can be opened in ArcView with the ECW extension from ERMapper.In the file names:ms=multispectral (2.4m)ps=pansharpened (0.6m)orc=orthorectifiedvis=visible colour composite (RGB, 8-bit: 0-255)fc=false colour composite (NIR,R,G 8-bit: 0-255)All images without fc or vis in the filename have four bands and have a 16-bit data type (0 - 65535). Quickbird collects image data in 11-bit (0-2048), so in order to display these images the image values need to be stretched. For GIS use Arko recommends using the fc and vis files, because they have been stretched already. Keep in mind that ArcGIS has to build image pyramids and image statistics before display.In 2012 lakes, creeks, escarpments and ridges were digitised from the orthorectifed image. For details about the digitising and access to the resulting dataset, refer to the metadata record 'Macquarie Island - Digitising topographic features', Entry ID: Macquarie_Island_quickbird_digitising.

2005年3月15日获取的麦夸里岛（Macquarie Island）快鸟（Quickbird）卫星影像。该影像已完成正射校正，即校正了所有由地球曲率与地形起伏引发的几何畸变。
基于测量控制点与该快鸟影像全色波段的对应像素，在影像覆盖范围内共识别出26个地面控制点。此项工作与澳大利亚南极数据中心的亨克·布罗尔斯马（Henk Brolsma）合作完成。
影像基于快鸟RPC系数（传感器平台参数）与地面控制点完成几何变换。本次校正使用了美国国家航空航天局（NASA）获取的5米分辨率AIRSAR数字高程模型（Digital Elevation Model，DEM，元数据记录：麦夸里岛AIRSAR DEM / 条目ID：macca_dem_gis），以校正影像的地形起伏畸变。DEM采用双线性重采样算法进行重采样，影像则采用最邻近邻域重采样算法以保留原始像素值。
水平定位精度预估为5米以内或2个像素以内，但部分区域的精度可能超出该范围。坐标系采用WGS84基准面的UTM 57S投影带。
本次正射校正由塔斯马尼亚大学地理与环境研究学院空间信息科学中心（CenSIS）的阿尔科·卢西耶博士（Dr Arko Lucieer）使用ENVI 4.3软件完成，校正后的新影像导出为GeoTIFF与JPEG2000（无损压缩）格式。
快鸟影像包含4个多光谱波段：蓝、绿、红与近红外波段，分辨率为2.4米（像素尺寸）。快鸟同时可获取分辨率0.6米的全色灰度影像。
本次数据集包含两种彩色合成影像：
1. 真彩色合成（RGB=波段3、2、1）：匹配人类视觉系统的色彩感知方式，与航空航拍照片视觉效果高度相似。
2. 假彩色合成（RGB=波段4、3、2）：该合成方案引入了近红外波段（波段4），以红色高亮显示植被。植被中的叶绿素对近红外波段能量具有高反射率，因此在影像中呈现为亮红色；被啃食或枯萎的植被则会失去红色色调。
本次提供两种影像格式：
1. GeoTIFF（.tif与.tfw）：该影像未压缩，可直接在ArcView 3.x或ArcMap中打开，无需加载扩展插件。
2. JPEG2000：该格式为高压缩比无损压缩格式，影像质量与GeoTIFF完全一致，但文件体积更小。JPEG2000格式影像需借助ERMapper公司的ECW扩展插件方可在ArcView中打开。
文件名中的缩写含义如下：
ms=多光谱影像（分辨率2.4米）
ps=全色锐化影像（分辨率0.6米）
orc=正射校正影像
vis=真彩色合成影像（RGB模式，8位色深：0-255）
fc=假彩色合成影像（近红外、红、绿波段组合，8位色深：0-255）
所有文件名中未包含fc或vis的影像均为4波段影像，数据类型为16位（取值范围0-65535）。快鸟卫星采集的原始影像数据为11位（取值范围0-2048），因此若需展示这些影像，需对影像像素值进行拉伸处理。阿尔科博士推荐在GIS应用中使用fc与vis格式的影像，因其已完成像素值拉伸。请注意，ArcGIS需先构建影像金字塔与影像统计信息后方可正常显示影像。
2012年，研究人员从正射校正后的影像中数字化提取了湖泊、溪流、悬崖与山脊要素。如需了解数字化工作的详细信息以及获取该衍生数据集，请参阅元数据记录《麦夸里岛——地形要素数字化》，条目ID：Macquarie_Island_quickbird_digitising。