Multiple technologies like light-field, laser imaging detection and ranging (LIDAR), infrared (IR) or stereo can also obtain the depth information, that is useful for a better understanding of the scene.
SABATO CERUSO & RICARDO OLIVA
Conventional cameras capture images from a scene just obtaining color information, resulting in a 2D picture. Multiple technologies like light-field, laser imaging detection and ranging (LIDAR), infrared (IR) or stereo can also obtain the depth information, that is useful for a better understanding of the scene.
What gets 3D devices besides conventional camera imaging?
Cameras sensor captures the light information of the scene, obtaining the wavelength in the visible range. The sensor reads each channel (RGB) separately, using a grid known as bayer pattern.
Bayer pattern grid, this grid is used to extract the RGB color of an image.
In order to obtain the depth information with a conventional camera an additional 3D sense technology is needed. Allowing to obtain the volumetric information of the scene.
Result of a conventional camera adding 3D information.
What kind of 3D technologies exists, and their differences?
There are two major categories in 3D capturing, active and passive methods. The first one uses light emission to obtain the information, waiting for the bounce of light to be captured (LIDAR, IR). Passive methods use only the obtained light by the sensor (stereo or light-field).
Different kinds of 3D technologies and their appearance.
While active methods work properly indoor obtaining high depth resolution and precise distances, but poorly in outdoor scenes; passive methods work in both. However, passive methods have resolution issues. Stereo based method loss depth resolution due to the distances between the two cameras. Light-field is more accurate in distances than stereo, but loses spatial resolution, as the size of the sensor is divided by the number of the micro lenses.
Usefulness of 3D capturing
Multiple applications become possible thanks to 3D information. All in focus image, defocus on demand, multiple viewpoints generation or create a 3D mesh to reconstruction of the scene are some. This kind of information is very useful for artist or filmmakers, allowing to make multiple changes in postproduction (like defocus), without the need of using different kind of camera objectives in the production process to generate a bokeh effect or to focus at different zones.