Abstract
The web has an enormous collection of live cameras that image parks, roads, cities, beaches, mountains, buildings, parking lots. There are a wide variety of problems that could effectively use this massively distributed, scalable, and already existing camera network. To move towards this goal, this paper discusses ongoing research with the AMOS (Archive of Many Outdoor Scenes) database, which includes images from 1000 cameras captured every half hour over the last 3 years. In particular, we offer (1) algorithms for geo-locating and calibrating these cameras just from image data, (2) a set of tools to annotate parts of the scene in view (e.g. ground plane, roads, sky, trees), and (3) advances in algorithms to automatically infer weather information (e.g. wind-speed, vapor pressure) from image data alone.