eduzhai > Applied Sciences > Engineering >

Single Image Brightening via Multi-Scale Exposure Fusion with Hybrid Learning

  • king
  • (0) Download
  • 20210505
  • Save

... pages left unread,continue reading

Document pages: 11 pages

Abstract: A small ISO and a small exposure time are usually used to capture an image inthe back or low light conditions which results in an image with negligiblemotion blur and small noise but look dark. In this paper, a single imagebrightening algorithm is introduced to brighten such an image. The proposedalgorithm includes a unique hybrid learning framework to generate two virtualimages with large exposure times. The virtual images are first generated viaintensity mapping functions (IMFs) which are computed using camera responsefunctions (CRFs) and this is a model-driven approach. Both the virtual imagesare then enhanced by using a data-driven approach, i.e. a residualconvolutional neural network to approach the ground truth images. Themodel-driven approach and the data-driven one compensate each other in theproposed hybrid learning framework. The final brightened image is obtained byfusing the original image and two virtual images via a multi-scale exposurefusion algorithm with properly defined weights. Experimental results show thatthe proposed brightening algorithm outperforms existing algorithms in terms ofthe MEF-SSIM metric.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×