Image Matting using Neural Networks

Image matting, also refers to picture matting in the article, is task of finding appealing targets a or sequence pictures i.e., video, and it has been used extensively many photo video editing applications. composition process extracting an eye-catching subject from photograph blending with different background. a) Blue/Green screen (curtain) where backdrop clear readily distinct between foreground (frontal area) background (foundation) portions. This approach now most type image matting. b) Natural which these sorts photos are taken naturally using cameras cell phones during everyday activities. These present known techniques It difficult discern distinction frontal area foundation at their boundaries. The current framework requires both RGB trimap images as inputs for natural compute since additional required obtain this trimap. study will introduce Picture Matting Neural Net (PMNN) framework, utilizes single input creates alpha matte without any human involvement user, overcome drawbacks prior frameworks. created tested against PPM-100 data set, PSNR SSIM measurement index utilized compare two. works well can be fed regular mobile reducing clarity image.