High-quality color image demosaicing method based on convolutional neural network

Abstract

The invention discloses a high-quality color image demosaicing method based on a convolutional neural network, and belongs to the field of image signal processing. According to the method, in the color image demosaicing process, under the condition that the calculated amount is not greatly increased, the self-similarity and redundant information of the image are fully utilized, a color image demosaicing network is designed, and the color image demosaicing network is used for representing end-to-end mapping from the mosaic image to the color image; a training set for training the designed colorimage demosaicing network is made; parameters required by color image demosaicing network training are configured; network parameters are updated by using a random gradient descent method through themade training set; and a target color image with any resolution is directly reconstructed by using the trained network model to obtain a high-quality color image. The method has the advantages of high reconstruction efficiency and high reconstructed color image quality. The image demosaicing method can be used for finishing image demosaicing tasks in all cameras based on the Bayer color filter array.

Description

technical field [0001] The invention relates to a method for demosaicing a color image, in particular to a method capable of quickly acquiring a high-quality color image, and belongs to the field of image signal processing. Background technique [0002] Single charge-coupled device (CCD) or complementary metal-oxide-semiconductor (CMOS) sensors and color filter arrays (CFA) are widely used in modern color digital cameras. The layout design of the CFA will vary depending on the camera type, the most commonly used Bayer filter array (Bayer CFA). Independent of the sensor type, each sensor element registers only one intensity value in the three colors R, G, B according to the specific CFA. A full-color image is reconstructed from the incomplete sampling of the CFA color channel output, that is, the complete RGB three-color combination of each pixel is reconstructed. This reconstruction process is a color image demosaic. Image demosaicing As the initial stage of acquiring colo...

AI Technical Summary

Problems solved by technology

[0005] In view of the fact that the existing convolutional neural network-based algorithms cannot make full use of the self-similarity and redundant information of natural images to reconstruct color images under certain computational constraints, when the local geometric information cannot be correctly inferred from local pixels, it is easy to cause problems such as Moore In many scenes, there are problems such as low quality of reconstructed images. The technical problem to be solved by a high-quality color image demosaic method based on convolutional neural network disclosed by the present invention is: on the premise of ensuring fast reconstruction speed Next, demosaicing the mosaic image based on the convolutional neural network can effectively use the self-similarity and redundancy of the mosaic image to obtain a high-quality color image, which has the advantage of high quality reconstructed image

Images