image denoising based on sparse representation in dea

Dea has been successfully applied to a host of different types of entities engaged in awide variety of activities in many contexts worldwide. this paper presents a method forutilizing data envelopment analysis (dea) with sparse input and output using clusteringconcepts. the approach is based on decomposition of an image into multiple semanticcomponents which have various image processing applications such as image denoising,super-resolution, enhancement and inpainting. in this paper, we present self-learningbased image decomposition framework based on sparse representation. the proposedframework first learns an over-complete dictionary from the high spatial frequency partsof the input image for reconstruction purposes. we perform unsupervised clustering onthe observed dictionary atoms which allows us to identify image-dependent componentswith similar context information. different from previous image processing works withsparse representation, proposed method does not need training images. we conduct theproposed method for denoising a single image. we validate the results by using anotherdictionary learning method called denoising by sparse coding based on douglas-rachfordalgorithm. visually comparison and psnr improvement of the proposed method showedits robustness.