Logo image
Real-time video alignment and fusion using feature detection on FPGA
Thesis   Open access

Real-time video alignment and fusion using feature detection on FPGA

Robert Haywood Taglang
Master of Science (M.S.), Drexel University
Jun 2017
DOI:
https://doi.org/10.17918/etd-7416
pdf
Taglang_Robert_20172.91 MBDownloadView

Abstract

Field programmable gate arrays Streaming video Computer Engineering
Video fusion functions as a way to combine the important or useful parts of two or more sequences of images. The scenario presented is the use of Laplacian fusion to produce a single video composed of the fields of view of two cameras whose areas of focus differ substantially. This is not a useful real-time strategy unless the fields can be aligned. This thesis presents a system for detecting features using an FPGA implementation of SURF (Speeded-Up Robust Features), and aligning video streams by applying a transform generated from the key features.

Metrics

47 File views/ downloads
34 Record Views

Details

Logo image