Laserfiche WebLink
2.4 Task 3 Satellite image preprocessing including georeferencing, corrections and calibration <br />Objective: <br />To prepare the acquired multidate Landsat imagery for its optimal use in the irrigated lands / crop <br />classification in the 2003 irrigated lands update in the Lower Arkansas River Basin. Similar to subtask <br />2.2, the satellite imagery will be processed in two phases. In phase 1, Landsat frames will be <br />georeferenced and subsequently calibrated using currently available DOQQs in the western part of the <br />basin. In phase 2, the remaining Landsat frames will be preprocessed as soon as the updated DOQQs in <br />Prowers County become available from USGS (Task 2.1). <br />Methods: <br />• Import imagery into ERDAS Imagine ( *.img format) <br />• Geometric Correction consisting of georeferencing the Landsat images to Digital Ortho <br />Quadrangles (DOQ). This procedure allows us to combine the high spatial accuracy of the DOQ <br />with the multi - spectral information of the Landsat imagery. <br />• Radiometric Correction consisting of two procedures: <br />- Image calibration. This procedure consists of the conversion of digital number (DN) <br />values to absolute radiance values and is required because each sensor (e.g., Landsat 7 <br />and Landsat 5) has its own calibration parameters used in recording the DN values. <br />- Atmospheric correction. In addition to the instrument calibration, it is necessary to <br />normalize image pixel values for differences in sun illumination geometry and <br />atmospheric effects. This improves the ability to mosaic adjacent imagery and compare <br />imagery over time. <br />• Image subset for study area. This procedure consists in `cutting' the satellite images to the extents <br />of the project area. This can reduces the amount of data handling and processing time <br />considerably. <br />Deliverables: <br />• Technical memorandum describing the procedures performed and deliverables <br />• CDs with the preprocessed Landsat image in ERDAS imagine ( *.img) format (complete <br />frames) <br />• CDs with the preprocessed Landsat imagery in ERDAS imagine ( *.img) format (subset to study <br />area) <br />• Metadata <br />3. Personnel <br />Timothy Martin, Ph.D. will serve as Principal -in- Charge. Claudio Schneider, Ph.D. will serve as Project <br />Manager and Remote Sensing Specialist. RTi also will provide the services of Nil Babel, GIS Specialist, <br />as well as the services of a GIS /RS Technician 11. <br />Page 3 of 4 <br />