Digital Image Processing

Vidya V. Shengule ¹, Rajesh K. Dhumal ¹, K. V. Kale ¹

Affiliations
1 Department of Computer Science and Information Technology, Dr. Babasaheb Ambedkar Marathwada University, Aurangabad, Maharashtra, IN

Abstract

In recent years, remote sensing data acquired through different satellite sensors .This acquired data has some geometric distortions. This raw data does not have any coordinate system. Image preprocessing is require for rectifying or removing geometric distortions .In this work we have used geometric transformation to remove geometric distortions. Geometric transformation is the process of using a set of control points and transformation equations to register a digitized map, or a satellite image onto a projected coordinate system. For geometric transformation, coordinate geometry is used. We have used satellite images such as (Thematic Mapper) TM, (Multispectral Scanning System) MSS, Resourcesat-I (Linear Imaging Self Scanner) LISS-III. To update existing map and to detect changes in remote sensing images, they must be reprojected to a coordinate system with known geometry identical to that of the digital maps to be revised. The order of the transformation (first-order, second-order, and third-order) affected the spatial accuracy of typical georectified satellite imagery. According to satellite image or pixel size, spatial distribution and measurement errors in a set of ground control points are determined. Accuracy of rectification is determined by Root-Mean-Square-Error (RMSE). RMSE is calculating the distance between the true locations and the transformed locations of the output control points. Rectification accuracy is small for low resolution and large for high resolution. This study demonstrates the rectifications of images of a coarser resolution are more accurately than for images of a fine resolution.

Keywords

Image Georeferencing, Image Rectification Model, Image Resampling, and Image Transformation.

Full Text