A NEW DEPTH RECOVERY ALGORITHM BASED ON DEFOCUS IMAGE

Cheng Hong,Zhang Quanbing,Gong Yanyan
DOI: https://doi.org/10.3969/j.issn.1000-386X.2010.02.087
2010-01-01
Abstract:To find the depth from defocus image is an important subject in computer vision.Since the blur level of a point in defocus image varies according to the depth of the object,thus defocus image can be used to estimate the depth information of an object.The approach proposed in this paper does not have matching problem of corresponding points in stereo vision and motion vision,so it has quite good applicative prospects.An algorithm of depth estimation based on defocus image space is studied.In the algorithm,the defocused imaging is expressed in heat diffusion,and two defocus images are expanded to a defocus space by means of warping function,then the warping parameters are estimated,followed with the recovery of the depth information of the object.Finally,with experimental results the validity of the algorithm is proved.
What problem does this paper attempt to address?