Photosynth

This week’s lecture was about multiplicity. We looked at the power of vast numbers on the internet and what increasing groups of things are capable of. One video we watched in the lecture was of a new software application by Microsoft called Photosynth. This is where multiple pictures of the same area are matched together and generates a three-dimensional model of the photos and a point cloud of a photographed object.

Photosynth technology works in two steps. The first involves the analysis of multiple photos taken of the same area. This process identifies specific features in a photo and then it is compared and matched to the same identified features in other the photos. By analysing the location of matching features with in each photo, the program can identify the 3D position of each feature in the photo as well as the position and angle each photo was taken. The second step involves the display and navigation through the 3D point cloud of images. This is with the downloadable photosynth viewer. It uses technology from the Seadragon program to zoom through high resolution photos without the user having to download them to their machine.

The man who spoke on the video about photosynth talked about how it could use Flickr as the source of photos and create detailed 3D models of the selected location. If the thousand people who visited these locations were to upload just one photo each, there would be a vast collection of images which could be used to create these 3D models.