What do you get if you cross a Kinect, a laptop, a car, and a smartphone? Well, if you’re Martin Szarski you get a hack that builds 3D maps of the real world. Using the Kinect’s motion sensor and depth mapping capabilities, Martin Szarski utilised the celebrated bedroom coding technique of capturing himself and his environment and took it for a spin in the outside world.
To account for the fact that he needed to know the exact coordinates of each Kinect snapshot to be able to stitch the photos together later, he cleverly used his Google Nexus One’s GPS to log the distinct real-time locations of each photo. By connecting the GPS and Kinect to his laptop, along with implementing some custom-built software, he was able to capture real-time 3D maps of any location, all with proper real world coordinates.
So how do the results compare to Google’s infamous all-seeing-100-megapixel-eye? Well, you can take a look above and judge for yourselves. This first implementation is fairly rudimentary, at best, but it’s still pretty impressive for a DIY effort. The first rendering is a twenty-six metre section of road and the second photo depicts a white picket fence.