Defocus Magnification

Abstract

b) defocus map (a) input (c) our result with magnified defocus Figure 1: Our technique magnifies defocus given a single image. Our defocus map characterizes blurriness at edges. This enables shallow depth of field effects by magnifying existing defocus. The input photo was taken by a Canon PowerShot A80, a point-and-shoot camera with a sensor size of 7.18 × 5.32 mm, and a 7.8 mm lens at f/2.8. Abstract A blurry background due to shallow depth of field is often desired for photographs such as portraits, but, unfortunately, small point-and-shoot cameras do not permit enough defocus because of the small diameter of their lenses. We present an image-processing technique that increases the defocus in an image to simulate the shallow depth of field of a lens with a larger aperture. Our technique estimates the spatially-varying amount of blur over the image, and then uses a simple image-based technique to increase defocus. We first estimate the size of the blur kernel at edges and then propagate this defocus measure over the image. Using our defocus map, we magnify the existing blurriness, which means that we blur blurry regions and keep sharp regions sharp. In contrast to more difficult problems such as depth from defocus, we do not require precise depth estimation and do not need to disambiguate textureless regions.

A focus fusion framework with anisotropic depth map smoothing

“…Mahmood and Choi [18] employ 3-D anisotropic diffusion to enhance the input images, and in turn, to obtain an accurate decision map. Staying with the idea of operating on decision maps, Bae et al [21] apply bilateral filtering to this decision map in a related context. However, they do not consider the focus fusion task, but perform a defocus magnification given a single image.…”

A Review of Geometry Recovery from a Single Image Focusing on Curved Object Reconstruction

“…However, the accuracy of such methods is limited and camera calibration is necessary. References include works from Levin [32] and Bae and Durand [1].…”

A Weighted Integral Method for Parametrically Describing Local Image Appearance

“…It has a scale parameter, of which the value can be estimated to describe the blurring scale at each location in given images. Such descriptions of the blurring scales are required for many applications, e.g., image matting [11], moving object detection [5], image enhancement [3], and 3D-shape reconstruction from defocus [6], [10]. Many methods, hence, have been proposed for estimating the blurring scale at each location in the given images.…”