A new sense for depth of field

TLDR: In this paper, the authors examined a novel source of depth information: focal gradients resulting from the limited depth of field inherent in most optical systems and proved that this source of information can be used to make reliable depth maps of useful accuracy with relatively minimal computation.

Abstract: One of the major unsolved problems in designing an autonomous agent [robot] that must function in a complex, moving environment is obtaining reliable, real-time depth information, preferably without the limitations of active scanners. Stereo remains computationally intensive and prone to severe errors, the use of motion information is still quite experimental, and autofocus schemes can measure depth at only one point at a time. We examine a novel source of depth information: focal gradients resulting from the limited depth of field inherent in most optical systems. We prove that this source of information can be used to make reliable depth maps of useful accuracy with relatively minimal computation. Experiments with realistic imagery show that measurement of these optical gradients can potentially provide depth information roughly comparable to stereo disparity or motion parallax, while avoiding image-to-image matching problems. A potentially real-time version of this algorithm is described.