Parallax is something nearly every human, and every animal with a set of forward facing eyes, utilizes every day and they may not realize it. Parallax is the shift in position of objects in a scene when there is a change in the position of the viewer (usually a camera or one of your eyes). Our brains use the parallax in the images from our eyes to create a sense of depth in our vision (stereopsis).
In the image at the beginning of the post we see an example of parallax at play. In one image the tree is centered in the image. As the photographer shifts their position, but is still centered on the tree the author can be seen hiding behind the tree. The object (the author) has apparently moved into the field of view of the photographer. I wonder what he is hiding from…
People have leveraged parallax in multi-camera imaging systems to mimic stereopsis to create real ‘3D’ images. The apparent separation of the cameras in these systems is important to simulate a sense of depth that is comfortable for a person to view. Mismatched parallax (from improper control of the effective Interpupillary Distance) of a two camera system will create images that have a ‘weird’ feeling of depth to users. If a system has small amounts of parallax images will appear to have no stereo depth, too much parallax and the field of view mismatch may causes a feeling of depth that is unnatural to a viewer.
Starting in the 1930’s many animated studios developed camera systems to image animation planes at multiple distances and would pan across the images. The parallax between the different image planes would give the scene a feeling of depth. Here is an informative video on how Disney used these multiplane cameras.
For other systems parallax is detrimental to the performance of the system. In World War II many complicated sighting systems using mechanical computers (such as the Norden bomb sight) were developed to overcome parallax issues bombardiers would experience while sighting targets. Optical sights and rangefinders can have parallax issues if the targeting system is used to view objects that are far from the ideal target distance (usually optical infinity)
Red-dot sights are often focused at long distances (optical infinity) and exhibit very little parallax when a user shifts their eye, but for close distances parallax effects can become noticeable. The optical path of the eye and the path of the ‘red-dot’ become different and parallax can be experienced.
In most LIDAR applications parallax can limit the useable ranging distance of the optical system. Often the illumination and detection channels are in separated optical paths. The ranging ability of these systems is limited to the overlap of the fields of view.
In all situations, the optical design of systems that can have parallax must be considered to either exploit or minimize the effects of the phenomena. Optikos has many experiences with designing systems that either leverage the benefits or mitigate the drawbacks of parallax. In future posts we will explore some of these design challenges.
Written by David P. Biss, Ph.D., Optikos Corporation