Super-Resolution Mosaicing of Unmanned Aircraft System (UAS) Surveillance Video Frames

Unmanned Aircraft Systems have been used in many military and civil applications, particularly surveillance. However, video frames are often blurry, noisy, and exhibit insufficient spatial resolution. This project aims to develop a vision-based algorithm to improve the quality of UAS video frames. This algorithm will be able to generate high resolution mosaic output through a combination of image mosaicing and super-resolution (SR) reconstruction techniques. The mosaicing algorithm is based on the Scale Invariant Feature Transform, Best Bins First, Random Sample Consensus, reprojection, and stitching algorithms. A regularized spatial domain-based SR algorithm is used to super resolve a mosaic input. The performance of the proposed system is evaluated using three metrics: Mean Square Error, Peak Signal-to-Noise Ratio, and Singular Value Decomposition-based measure. Evaluation has been performed using 36 test sequences from three categories: images of 2D surfaces, images of outdoor 3D scenes, and airborne images from an Unmanned Aerial Vehicle. Exhaustive testing has shown that the proposed SR mosaicing algorithm is effective in UAS applications because of its relative computational simplicity and robustness.