Building off the technology in the Microsoft Kinect, a team at Northwestern University has developed the fastest, highest quality 3D camera to date. After the Kinect was released in 2010, Microsoft made the 3D capturing technology available for other uses, and to date that technology has been put to use in robotics and other fields. The problem, though, is that the Kinect is not the most reliable or accurate of cameras, so its use is limited.
For example, the Kinect doesn’t work outside because sunlight overpowers the light the camera uses, making it impossible to capture images. Furthermore, it’s not terribly precise, since it sends light out to capture motion in a scene by processing light in every pixel of the camera at once. This works for short distances, where movement is the only real concern, but it’s not as useful in the real world.
Compare this to single-point scanners, which only process one pixel at a time. They’re more precise, but they take much longer.
The new camera developed by Oliver Cossairt and his team at the Northwestern University McCormick School of Engineering uses a single-point scanner, but it reduces time by only scanning new information instead of constantly rescanning the entire scene.
Cossairt compared it to the function of the human eye, which ceases to process a scene if nothing changes, saving on “processing power” in the brain. This camera works in much the same way, allowing the camera to return precise data faster and more efficiently. The laser they use is also bright enough to be used outside without the sun overpowering it.
The camera has a lot of potential uses, such as robotics, automation, or bioinformatics. The team received a Google Faculty Research Award, which they are using to work the camera into an autonomous vehicle design. The goal is to integrate the camera in such a way that it will allow the vehicle to “see” the area in front of it and move acco