 
Summary: GSV Accomplishments
Preliminary Assessment Results
and Gap Analysis
Proposed Validation Process Detailed Description of Proposed V&V Procedure
· "On the verification and validation of geospatial image analysis algorithms"
accepted for publication at the IEEE 2010 International Geoscience and Remote
Sensing Symposium
· Preliminary annotated bibliography of papers relevant to GSV technologies
· Support SAM ontology effort with respect to relevant geospatial concepts,objects
and relationships
· Preliminary design of a V&V methodology for GS algorithms
· Determine test parameters such as:
Metric to compare response of ideal algorithm to test algorithm: M(R,Rideal)
Criteria for success of algorithm and test
· Begin process with N algorithms (ALGn n = 1 to N) and K images (IMGk k = 1 to K) to test
the algorithms
· Loop over all K images for nth algorithm,recording response to each image
Mnk(Rnk,Rideal) {M}n. Determine whether algorithm meets success criteria:
Algorithm is considered validated if it meets success criteria
Option to algorithm developer to modify algorithm if it doesn't meet success
