Images are an integral part of content transmitted in info-communication systems. 

Traditional statistical image filtering algorithms are not always effective for the random nature of the noise spectrum. 

Image denoising by convolutional neural networks is a modern and effective approach. 

The article is devoted to demonstrate the possibilities of using denoising convolutional neural networks to solve one of the most difficult tasks that developers face when performing the transfer of graphical information in infocommunication systems - denoising. 

Unlike traditional algorithms, denoising convolutional neural networks have architectural features that allow them to perform effective image filtering with an unknown noise level. 

It is suggested to use denoising convolutional neural networks to generate a correction signal in the info-communication system, which transmits a noisy image.