A Fine Summer Day in 1951: Creating a Colorized 3D Model of Washington, DC Using Historical and Modern Aerial Imagery and LiDAR

Authors: Zachary Bortolot*, James Madison University
Topics: Remote Sensing, Historical Geography, Land Use and Land Cover Change
Keywords: Washington DC, historical aerial imagery, visualization, LiDAR
Session Type: Paper
Day: 4/5/2019
Start / End Time: 5:00 PM / 6:40 PM
Room: Roosevelt 7, Marriott, Exhibition Level
Presentation File: No File Uploaded

Historical aerial imagery is a rich source of information about the past, and it has the potential to help citizens understand and appreciate change so they can make more informed decisions in the future. Unfortunately, most historical imagery is panchromatic and presented to the audience as 2D images rather than 3D models like those available on Google Earth and other online mapping platforms, and this limits the impact of the images and people’s ability to interpret them. For this project, a colorized 3D model of Washington, DC was produced from 1:8,000 scale panchromatic aerial imagery collected on July 5, 1951. The major steps of the process were: 1. Orthorectify and mosaic the imagery to produce an orthoimage with a 25cm resolution; 2. Extract a digital surface model (DSM) from the imagery photogrammetrically; 3. Merge the historical orthoimage with a modern image and perform a texture-based change detection to identify changed pixels; 4. Perform an automated colorization process that involves taking the color information from the modern color image for pixels that had stayed the same, and obtaining color information from similar features for changed pixels; 5. Perform a segment-based color filtering procedure; 6. Use a LiDAR-derived DSM to make the historical DSM more accurate in areas that had not changed; and 7. Create 3D models using the colorized image and historical DSM. The results show that the colorization was realistic and the model shows elevations accurately. A sample can be found at this link: https://tinyurl.com/DC-3D-Demo

Abstract Information

This abstract is already part of a session. View the session here.

To access contact information login