Poisson matting

By Jian Sun, Jiaya Jia, Chi-Keung Tang, Harry Shum. ACM Transactions on Graphics, 2004

480 citations recorded by the source corpus.

In this paper, we formulate the problem of natural image matting as one of solving Poisson equations with the matte gradient field. Our approach, which we call Poisson matting, has the following advantages. First, the matte is directly reconstructed from a continuous matte gradient field by solving Poisson equations using boundary information from a user-supplied trimap. Second, by interactively manipulating the matte gradient field using a number of filtering tools, the user can further improve Poisson matting results locally until he or she is satisfied. The modified local result is seamlessly integrated into the final result. Experiments on many complex natural images demonstrate that Poisson matting can generate good matting results that are not possible using existing matting techniques.

Subfields: Computer Graphics and Computer-Aided Design, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/77740690e2423078279b18c925cc8712bfbf64a1

Citations in the corpus

• A Geodesic Framework for Fast Interactive Image and Video Segmentation and Matting (2007, unknown): “Poisson matting, [15], and Bayesian matting, [7], are two important matting techniques that use trimaps as inputs.”
• A perceptually motivated online benchmark for image matting (2009, unknown): “1 For all algorithms, we used the implementations of the respective authors, except for [20] which we implemented ourselves.”
• Lucid Data Dreaming for Object Tracking (2017, unknown): “Fg/Bg merge: finally (Iτ−1, Iτ ) are composed by blending the perturbed foreground with the perturbed background using Poisson matting [49].”
• An iterative optimization approach for unified image segmentation and matting (2005, unknown): “The most successful systems include Knockout 2 [3], the approach proposed by Ruzon and Tomasi [7], Bayesian matting [12] and Poisson matting [5].”
• Real-Time High-Resolution Background Matting (2020, unknown): “Matting with known natural background had been previously explored in [23], Bayesian matting [7] and Poisson matting [29, 10] which also requires a trimap.”
• A Closed-Form Solution to Natural Image Matting (2006, unknown): “Most existing methods for natural image matting require the input image to be accompanied by a trimap [1], [2], [5], [6], [14], [17], labeling each pixel as foreground, background , or unknown.”
• Removing photography artifacts using gradient projection and flash-exposure sampling (2005, unknown): “Our gradient domain techniques are related to those used for HDR compression [Fattal et al. 2002], image editing [Perez et al. 2003], image fusion for context enhancement [Raskar et al. 2004b], creating photomontages [Agarwala et al. 2004] and image matting [Sun et al. 2004].”
• Local color transfer via probabilistic segmentation by expectation-maximization (2005, unknown): “For probabilistic segmentation, note that natural matting techniques (such as [2] and [11]) are not applicable in our work, since they do not guarantee smoothness for in certain situations.”
• Guided Image Filtering (2010, unknown): “LTI filtering also includes the process of solving a Poisson Equation, such as in high dynamic range (HDR) compression [4], image stitching [5], and image matting [6], where the filtering kernel is implicitly defined by the inverse of a homogenous Laplacian matrix.”
• Single Image Motion Deblurring Using Transparency (2007, unknown): “We do not propose a new method to compute the transparency since there have been several successful natural image matting techniques proposed [3, 22, 14] which estimate foreground, background, and the alpha matte from a single input natural image.”
• A global sampling method for alpha matting (2011, unknown): “The interpolation can be done by solving an affinity matrix [17, 8, 11, 16, 9], by optimizing Markov Random Fields [18], or by computing geodesic distance [2].”
• Fast, automatic and fine-grained tampered JPEG image detection via DCT coefficient analysis (2009, unknown): “The trustworthiness of photos is being undermined by the advance of image/video editing techniques (e.g., [1] [2] [3] [4] [5] [6] [7], to name just a few).”
• Geodesic Matting: A Framework for Fast Interactive Image and Video Segmentation and Matting (2009, unknown): “Poisson matting, [24], and Bayesian matting, [11], are two important matting techniques that use trimaps as inputs.”