Advancements in digital photography and facilities for storing and sharing photos on the Internet have made digital photography accessible to millions of people. This article shares the authors’ work with online photo collections to recreate information about the world and its people, using automatic algorithms that analyze photos and associated metadata such as geotags, timestamps, captions, user profiles, social contacts, and so on. These algorithms succeed at automatically identifying most photographed places; creating 3D models from 2D photographs; and discovering people’s travel patterns, among other similar tasks.

The authors explain how data is cleaned, how useful photos are separated from those that lack relevant information, how metadata is obtained and used, how their algorithms work, and the principles and concepts used in developing those algorithms. They also explain the limitations of available photo collections and describe techniques, such as using computer games, to motivate people to take photos that will help in recreating the world. The article concludes with a discussion of the potential uses of this technology in other fields and the need for future work to make the system foolproof and efficient.

This is a must-read for people working in computer vision, pattern recognition, data mining, knowledge discovery, artificial intelligence, and related fields, as it opens a new avenue for gathering information from photo collections. This would also help scientists, engineers, meteorologists, cartographers, and professionals in related fields. The ability to create 3D models from 2D photographs and recreate the sequence of events at a particular place from photos taken over a period of time can be very useful as an investigation tool.