Authors: Kangsan Lee*, University of Iowa, Marc Linderman, University of Iowa
Topics: Remote Sensing, Environment
Keywords: UAS, Image Matching, Contrast Adjustment, Image Filtering
Session Type: Poster
Start / End Time: 1:20 PM / 3:00 PM
Room: Napoleon Foyer/Common St. Corridor, Sheraton, 3rd Floor
Presentation File: No File Uploaded
Several stereo image matching algorithms have been introduced to improve point matching for Structure From Motion image analysis. These algorithms are based on finding scale, rotation, and brightness invariant points to align images. Although these algorithms are a powerful solution to process point matching in general, sometimes they have limitations when using images with low texture or contrast image datasets due to a lack of sufficient tie points and their descriptions. This research focuses on the global and local scale preprocessing to enhance the original images obtained by a UAV in off-season agricultural fields in Iowa. It has been observed that global scale preprocessing increases tie point matching, while the adaptive preprocessing increases the quality of point descriptors.