MSDI: a geo-tagged drone imagery for absolute visual localization

The MSDI (Manchester Surface Drone Imagery) is a geo-imagery registration dataset. <br> The dataset consists of <br> a. 446 downward-facing drone images.<br> b. 89 forward-facing(45-degree) drone images.<br> c. 64 forward-facing (0-degree) drone images. <br> d. parameter matrix of drone camera and transformation matrix.<br> e. checkerboard images for camera calibration. This dataset is collected by Mochuan Zhan for his MSC project: Registration of UAV Imagery to Aerial and Satellite Imagery<br> in the University of Manchester (2021/9 - 2022/9) which is supervised by Dr.Terence Patrick Morley. This project aims at <br> developing a system that could perform efficient UAV visual localization through image registration based on local feature <br> detectors and the technique of high-throughput computing. Notice:<br> The corresponding satellite image from Google Map and Bing Map could be obtained by my program, Link: <pre>https://doi.org/10.5281/zenodo.6977652</pre> A Ground Control Point (GCP) selector is provided for user to select GCP and create file with small effort.<br> By registrating corresponding images, users could evaluate the performance of their registration techniques. The Imagery contains images of 8 Areas in Manchester:<br> - Manchester Aquatics Center 80<br> - Manchester ASDA 76<br> - Manchester Bussiness School 37<br> - Manchester Energy Center 71<br> - Manchester Holy Name Church 58<br> - Manchester Hulme Park 47<br> - Manchester Hulme Part(0-degree) 64<br> - Manchester Hulme Part(45-degree) 89<br> - Manchester Metropolitan university 29 <br> - Manchester Museum 48 Device Information:<br> - Drone brand: Parrot<br> - Drone model: Parrot Anafi Software Information:<br> - Pix4DCapture<br> - FreeFlight6 Flight parameters:<br> - Height 100m<br> - Speed 5m/s<br> - overlap low <br>