How would one create a map for gaming using Google street view? Are there any programs for doing this? The cars, people and objects could be edited out and replaced with digital copies if the same objects for that area. Originally I was thinking GTA Local but would require a much more work.

3D game with Pre-rendered background


enter image description here

  • \$\begingroup\$ This is an area of active research for machine learning and computer vision applications. You're unlikely to find a concise enough solution to fit in a StackExchange answer though. Also, don't forget to verify whether this is allowed under Google's licensing of the StreetView data. \$\endgroup\$
    – DMGregory
    Mar 3, 2019 at 23:20
  • \$\begingroup\$ Vote to close as too broad. Hi @Muze, your question seems to be asking for recommendation of software, which is usually off-topic here. The question could be more on-topic if it had the form "I tried doing X, but failed, what did I do wrong?", but asking about how to turn Google Street View into a video game, is way too broad, and touches plenty of complicated fields. \$\endgroup\$ Mar 4, 2019 at 15:39
  • \$\begingroup\$ This would make a great research topic for a group of 5 fully-funded PhDs looking to spend 3 years on it. (Voting to close as too broad.) \$\endgroup\$
    – Anko
    Mar 6, 2019 at 5:06

2 Answers 2


This is, in a way, available from Google via their Google Maps Platform. The data provided by this platform is a combination of aerial, satellite and street view data.

I'm not sure how far down the thought process of your idea:

The cars, people and objects could be edited out and replaced with digital copies if the same objects for that area

But that would certainly be exceptionally difficult with just the street view data alone (even more so with the limited data provided to end users). Utilizing the above listed combination of data sources, Google is able to use complicated machine learning and computer vision AI to remove the people, cars and other dynamic objects in a city environment.


What you would need to do is:

  1. detect and remove certain kinds of objects from static photos
  2. generate 3d models of environments from static photos

and those fully automatized and in real-time.

Doing things like that is an area which is currently heavily researched. The usual methods people use to solve them are machine learning and other AI based approaches. While I have seen some pretty impressive proofs of concept in these areas, I doubt that the technology is already far enough to do what you want. And most of what I have seen are proprietary technologies not available to the public, at least not for a license fee anyone can afford.

My estimate is that you should give the technology about ten more years.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .