Week 8 Final Exam Study Guide NR 503 - Updated for Current Exam
Week 8 Final Exam Study Guide NR 503 - Updated for Current Exam Workflow Components Post Download - answer - Stack, Scale, Reproject, Georeference, Clip Geoprocessing Steps - answer - Check the coordinate systems of your project Put all of our objets in the same coordinate system and datum, and the same Projected Coordinate System. Coordinate Systems - answer - Areas are flattend around an origin to calculate distanes. Precision decreases as you get farther Unprojected Coordinates - answer - Lat Long that focues on angular geometry, UTM Zones - answer - Break the Map Area up into vertical slices Reproject - answer - Process where you get all your data into the same coordinate system that represents your study area closely. Orthorectification - answer - Topographic Normalization. Noramlizes variation in view and illumination angles. Requires good DEM models Georefernecing - answer - Spatially aligning things on the landscape, removing orbital distortion. Done by taking an object that is crisp in space and easily seen in all images. Resampling - answer - Nearest Neighbor Bilinear Interpolation Cubic Convolution Nearest Niehgbor - answer - Assigns values of closet pixel to assign output Good for perfectly nested pixels (20 - 10 pixel size) Bilinear Interpolation - answer - Best for when grids are misaligned, Reflectance Indices - answer - Simple Ratio and Normalized Differences Conduct Classification - answer - Unsupervised or Supervised, Cutoffs. Unsupervised Classification - answer - Cluster-based analysis informed by spatial and spectral patterns. Pattern seeking brain essentially Supervised Classification - answer - utilizes a model built off ground truth data and known cases Object Oriented Classificaiation - answer - Groups pixels based on spatial / spectral characteristics Information Classes - answer - Usually assign values based on spatial feature (Coarseness, direction, etc). Creates categories of interest to assign spetral classes to. e spectral classes - answer - groups of pixels that are uniform and close to eachother. Uses land cover context. Spectral Subclasstes - answer - different categories of the same main class to aount for factors like mountain facing Supervised Classification - answer - Creates training sets with formal knowledge from spectral signatures. Uses ground truth data and accuracy assessments Classification, Rule parallelepiped, Minmum Distance, Maximum, Likelihood, Spectral Angle mapper, etc. k means - answer - Specifying a number of clusters and values associated with them, Advantages of Unsupervised Classfiication - answer - No extesnive prior knowledge needed. Opportunity for human error minimzed Reveals unique spectral classes Used with faormal ancillary data and prope accuracy assesment Disadvantages of Unsupervised Classifications - answer - Relies of natural groupings, makes spectral classes Analyst hasl ittle control over image classes/clusters Specral properties hange over time Detailed spectral knonwledge of surfaces is necessary. Supervised Process - answer - analyst provides ground truth data Limited by quantiy and quality of ata. Ground truth data is broken tnto trianing and testing / validation for accuracy
Información del documento
- Subido en
- 4 de noviembre de 2026
- Número de páginas
- 3
- Escrito en
- 2026/2027
- Tipo
- Examen
- Contiene
- Preguntas y respuestas
