Remote sensing image performance metrics: Comparing ground sample distance and the national imagery interpretability rating scale

Two remote sensing image quality metrics, ground sample distance (GSD) and the recently declassified, national imagery interpretability rating scale (NIIRS) are compared. GSD measures the geometric projection of a single detector on the ground, not actual image interpretability. High resolution imagery is generated to test how imagery with a constant GSD deteriorates when key components of an imaging system are altered. The NIIRS rates the information potential of an image, i.e., what tasks could be accomplished with a particular NIIRS rating. Remote sensing imagery from six sensors taken of one location are compared for interpretability. Each is rated as part of the study, by a certified image analyst and by a modeling tool developed to compute image quality based on the two-dimensional, spatial frequency power spectrum. Results indicate that the model performs well and NIIRS ratings are more indicative than GSD rating of the potential interpretability of remote sensing imagery.