PROJECT TITLE :

A Novel Retinex Based Enhancement Algorithm Considering Noise - 2017

ABSTRACT:

The ancient center/surround Retinex enhancement algorithms solely take into account the illuminance. Therefore it may amplify the noise. However image details can be lost in the next denoising method. During this paper, we tend to propose an illuminance-reflectance model(IRMNE) which can denoise the image taken in the low lightweight and enhance it at the same time. IRMNE can overcome the restrictions of traditional algorithms. The experiment results indicate the validity and advantage of our algorithm. Compared with different algorithms, IRMNE makes a better balance between image enhancement and image denoising, which helps to urge a lot of reliable color restoration results.