Depth map is an image that encodes the information relating to the distances of the surfaces of scene objects from a viewpoint. Basically, it stores the distance information for every pixel of the image from the given viewpoint. It has a lot of uses in computer vision ranging from object segmentation to 3D structure computation.

LRSE, low resolution stereo estimator, is a solution we developed for estimating a 'quick and dirty' depth map for very low resolution image pairs. It requires very little memory and computational resources. It also fills-in the occluded regions by using the nearby depth content.

Below we provide some examples of the estimated depth maps for image pairs of 427x370 resolution. For more examples, visit our LowRes Stereo Estimation Gallery.

ex ex
ex ex
ex ex

Example color and depth images for LowRes Stereo