Improving alpha matting and motion blurred foreground estimation

Abstract

We present a new method for separating motion blurred foreground objects from their background given a single image. Previous techniques focused on estimating alpha mattes for separating sharp, non-moving foreground objects from fairly homogeneous background. In those cases the only pixels which are ambiguous are those which exhibit fractional pixel occupancy. In this paper, we address the problem of alpha matte and foreground estimation of motion blurred objects. We show, that explicit modeling of the object motion facilitates the estimation and improves the quality of the estimated alpha mattes. In addition, we improve foreground extraction of motion blurred objects with a new regularization term. This task is particularly difficult in smeared out regions, where the background shimmers through. Both synthetic and real-world examples illustrate the merit of our approach.

Semantic Image Matting

“…For works targeting at other specific classes of matting, Lin et al [27] introduced motion regularization for matting motion blurred moving objects. Köhler et al [22] proposed to separate motion blurred foreground through explicit modeling of object motion. Amin et al [4] applied image matting to segment out-of-focus regions.…”

Blur Feature Extraction Plus Automatic KNN Matting: A Novel Two Stage Blur Region Detection Method for Local Motion Blurred Images

“…Dai and Wu [15] regarded local blur images as two layers of foreground and background according to clear and blur areas and achieved the detection of blur areas by combining image matting technology, user mark and appropriate image prior model. Kohler et al [16] used the existing blind deconvolution method to estimate blur kernel for the coarse blur region manually calibrated by the users, and then guided the refinement of the blur region by combining the estimated blur kernel information with the lazy matting algorithm. However, most existing methods often require additional information such as user interaction or prior knowledge to achieve satisfactory detection performance.…”

Automatically Extract the Semi-transparent Motion-blurred Hand from a Single Image

“…However these methods cant deal with motion-blurred hands, which are very common in practical applications. Existing traditional methods [3][4][5], which were designed to predict the alpha mattes or foreground images of motion-blurred objects, generally need user interactions [3,4] or short-exposure frames [5]. Zhao et.…”