METHOD FOR ENHANCING IMAGE USING STRUCTURAL TENSOR BASED ON DEEP LEARNING AND DEVICE THEREOF

摘要

According to embodiments of the present invention, provided is a method for enhancing an image of a device for enhancing an image which comprises a processor, a memory storing a program executed by the processor, and an image reception unit receiving an input image photographed by an image photographing device. The method for enhancing an image comprises the steps of: allowing a processor to receive an input image including a channel image with regard to each of a plurality of channels; allowing the processor to extract a plurality of first characteristic maps for enhancing the image by using an input residual block (Resblock) performing convolutional calculation and filtering of the channel image; allowing the processor to generate an integrated characteristic map by coupling the plurality of extracted first characteristic maps to each other; and allowing the processor to generate a second characteristic map by using an output Resblock performing convolutional calculation and filtering of the integrated characteristic map and generate an output image in which distortion of the input image is improved through the second characteristic map.;COPYRIGHT KIPO 2020