Cameras in the Lab
RGB-D camera sensor combines information about colour and depth. This allows capturing texture from colour image and geometry of an object in the scene from depth sensor.
Some types of RGB-D cameras are:
- Stereo cameras: imitates human binocular vision because uses two or more lenses with a separate image sensor. So, it gives the ability to capture 3D images.
- Time-of-flight (ToF): measures the time that a ray or pulse of light takes to travel certain distance between the camera and the object. The entire scene is captured with single pulse of light, as opposed to point-by-point with a laser beam. ToF camera has an active light source that is used to illuminate the scene. The camera registers the light reflected from the scene and depth information is calculated. KinectOne uses time-of-flight technique.
- Structured light: is a technique to measure 3D surfaces by the projection of light patterns of infrared light. This patter is seen distorted when looked from a perspective different from the projector. By analysing this distortion, information about the depth can be retrieved and the surface reconstructed. The main advantage is that the projected features are easily distinguished by the camera. Kinect360 uses structured light technique.
Kinect360 and KinectOne are sensors for videogames developed by Microsoft. These cameras are widely used for the robotics community in the field of computer vision for its low cost.