Low-Light Enhancement in the Frequency Domain
This work addresses low-light image enhancement to improve visibility and reduce noise for applications like object detection and tracking, representing a novel method for a known bottleneck.
The paper tackled low-light image enhancement by proposing a residual recurrent multi-wavelet convolutional neural network (R2-MWCNN) that operates in the frequency domain, which outperformed state-of-the-art methods in both quantitative and qualitative metrics.
Decreased visibility, intensive noise, and biased color are the common problems existing in low-light images. These visual disturbances further reduce the performance of high-level vision tasks, such as object detection, and tracking. To address this issue, some image enhancement methods have been proposed to increase the image contrast. However, most of them are implemented only in the spatial domain, which can be severely influenced by noise signals while enhancing. Hence, in this work, we propose a novel residual recurrent multi-wavelet convolutional neural network R2-MWCNN learned in the frequency domain that can simultaneously increase the image contrast and reduce noise signals well. This end-to-end trainable network utilizes a multi-level discrete wavelet transform to divide input feature maps into distinct frequencies, resulting in a better denoise impact. A channel-wise loss function is proposed to correct the color distortion for more realistic results. Extensive experiments demonstrate that our proposed R2-MWCNN outperforms the state-of-the-art methods quantitively and qualitatively.