Unsupervised end-to-end demosaicing method and system

The invention discloses an unsupervised end-to-end demosaicing method and system, and relates to the technical field of image processing, and the method comprises the steps: collecting video images in different scenes through camera equipment, obtaining original image data from the video images, and constructing an initial data set; preprocessing the initial data set to obtain an enhanced training data set; inputting the training data set into the demosaicing model for network training until the loss function converges to obtain a trained demosaicing model; and processing the mosaic image by using the trained demosaicing model, and outputting a reconstructed color image. According to the technical scheme of the invention, the method can be better realized, is resistant to pseudo-color, and retains texture details, thereby generating a better detail effect. GT manufacturing is not needed, and the network complexity is low. 本发明公开了一种无监督的端到端去马赛克方法及系统，涉及图像处理技术领域，包括：利用摄像设备采集不同场景中的视频影像，从视频影像中获取原始图像数据，构建初始数据集；对初始数据集进