Exposure Correction Model to Enhance Image Quality

Exposure errors in an image cause a degradation in the contrast and low visibility in the content. In this paper, we address this problem and propose an end-to-end expo-sure correction model in order to handle both under- and overexposure errors with a single model. Our model contains an image encoder, consecutive residual blocks, and image decoder to synthesize the corrected image. We utilize perceptual loss, feature matching loss, and multi-scale discriminator to increase the quality of the generated image as well as to make the training more stable. The experimental results indicate the effectiveness of proposed model. We achieve the state-of-the-art result on a large-scale exposure dataset. Besides, we investigate the effect of exposure set-ting of the image on the portrait matting task. We find that under- and overexposed images cause severe degradation in the performance of the portrait matting models. We show that after applying exposure correction with the proposed model, the portrait matting quality increases significantly. https://github.com/yamand16/ExposureCorrection.

1. Introduction

The quality of the images depends on several factors and dramatically affects the performance of the computer vision methods. The exposure attribute is one of these factors and depends on shutter speed, f-number, and camera ISO. The exposure setting is expressed by exposure values (EVs) and each EV yields a different level of brightness in the image. In the case that zero EV value is the proper setting for an arbitrary image, negative EV makes it underexposed, while positive EV causes overexposed version. Besides, underexposed images have a darker appearance and overexposed images have a brighter view. Moreover, both situations cause low visibility. Therefore, exposure correction is a key step to overcome exposure errors to provide a better image.