Best (or any) way to programmatically screen grab a Google Street View panorama

Question

As the question implies, I'm looking for a way to programmatically screen grab a given panorama, I.E set a longitude and latitude and POV (pitch, yaw and zoom) and save the grab to the server. So far the most promise has been shown by

1. Using .net to control the google earth com api (http://earth.google.com/comapi/index.html), however I am unable to find a definitive answer on whether on not the street view layer is accessible via this means.

2. Embed the street view swf inside another swf that opens a socket to the web server to listen for requests, and passing commands (such as adjusting lat/lng and POV) to the street view swf and screen grabbing the view and saving it back to the server. The swf would either be running on the webserver or another server.

Questions about reliability and scalability come into play with both of these solutions. Has anyone got any further suggestions or ideas? The solution doesn't have to be real time, its assumed that some asynchronous "behind the scenes" processing will be happening.

Answer 1

Are you concerned with licensing? I would suspect that google limits what you can do with StreetView data, and likely prohibits "screen scraping"...

Assuming that you figured out how not to get sued by google, one option might be to trace the HTTP requests generated by the street view SWF, and try to reverse engineer the API. I would try to use http://fiddler2.com first (it is a marvellous tool for browser-based HTTP debugging), and if that does not work too well with the flash player, resort to http://www.wireshark.org (this is a general-purpose network sniffer, and might require some getting used to)