Targeting accurate object extraction from an image a comprehensive study of natural image matting
- 1. Logic Mind Technologies Vijayangar (Near Maruthi Medicals), Bangalore-40 Ph: 8123668124 // 8123668066 Title: Targeting Accurate Object Extraction From an Image: A Comprehensive Study of Natural Image Matting Abstract—With the development of digital multimedia technologies, image matting has gained increasing interests from both academic and industrial communities. The purpose of image matting is to precisely extract the foreground objects with arbitrary shapes from an image or a video frame for further editing. It is generally known that image matting is inherently an ill-posed problem because we need to output three images out of only one input image. In this paper, we provide a comprehensive survey of the existing image matting algorithms and evaluate their performance. In addition to the blue screen matting, we systematically divide all existing natural image matting methods into four categories: 1) color sampling- based; 2) propagation-based; 3) combination of sampling-based and propagation- based; and 4) learning-based approaches. Samplingbased methods assume that the foreground and backgroundcolors of an unknown pixel can be explicitly estimated by examining nearby pixels. Propagation-based methods are instead based on the assumption that foreground and background colors are locally smooth. Learning- based methods treat the matting process as a supervised or semisupervised learning problem. Via the learning process, users can construct a linear or nonlinear model between the alpha mattes and the image colors using atraining set to estimate the alpha matte of an unknown pixel without any assumption about the characteristics of the testing image. With three benchmark data sets, the various matting
- 2. algorithms are evaluated and compared using several metrics to demonstrate the strengths and weaknesses of each methodboth quantitatively and qualitatively. Finally, we conclude this Manuscript received December 8, 2013; revised May 20, 2014 and October 29, 2014; accepted November 3, 2014. Date of publication November 20, 2014; date of current version January 15, 2015. This work wassupported in part by the National Natural Science Foundation of China under Grant 61303166 and Grant 61125106, in part by the Key Research Program through the Chinese Academy of Sciences, Shenzhen, China, under Grant KGZD- EW-T03, and in part by the National Basic Research Program (973 Program) of China under Grant 2010CB732606. (Corresponding authors: Ling Shao, Lei Wang.) Q. Zhu is with the Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen 518055, China, and also with the ChineseUniversity of Hong Kong, Hong Kong (e-mail: qs.zhu@siat.ac.cn).L. Shao is with the Department of Computer Science and DigitalTechnologies, Northumbria University, Newcastle upon Tyne NE1 8ST, U.K. (e-mail: ling.shao@ieee.org). X. Li is with the Centre for OPTical Imagery Analysis and Learning, State Key Laboratory of Transient Optics and Photonics, Xi’an Institute of Optics and Precision Mechanics, Chinese Academy of Sciences, Xi’an 710119, China(e-mail: xuelong_li@opt.ac.cn). L. Wang is with the Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen 518055, China (e-mail: lei.wang@siat.ac.cn). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/TNNLS.2014.2369426 paper by outlining the research trends and suggesting a number of promising directions for future development.
- 3. Software & Hardware requirement: 1. Hardware requirement: 1. PC 2. RAM minimum 2GB 3. HDD minimum 100GB 2. Software requirement: 1. MATLAB 7.0 2. Signal processing toolbox 3. Image processing toolbox 4. Mathematical toolbox PROJECT FLOW: First Review: Literature Survey Paper Explanation Design of Project Project Enhancement explanation Second Review: Implementing 40% of Base Paper Third Review Implementing Remaining 60% of Base Paper with Future Enhancement (Modification) For More Details please contact Logic Mind Technologies Vijayangar (Near Maruthi Medicals), Bangalore-40 Ph: 8123668124 // 8123668066 Mail: logicmindtech@gmail.com