Blind denoising of dental X-Ray images

The present study addresses the issue of automatic analysis and noise reduction in dental X-ray images obtained through the Morita system. These images are characterized by spatially correlated noise with an unknown spectrum and varying standard deviation across different regions of the image. To address this issue, we propose the utilization of two deep convolutional neural networks. The first network estimates the spectrum and level of noise for each pixel of a noisy image, predicting maps of noise standard deviation for three different image scales. The second network utilizes these maps as inputs to suppress noise in the image. Results obtained using both modeled and real-life images demonstrate that the proposed networks achieve a peak signal-to-noise ratio (PSNR) for dental X-ray images that is 2.7 dB better than the state-of-the-art denoising methods.