Weak light color imaging method based on deep learning

Abstract

The invention discloses a weak light color imaging method based on deep learning. The method comprises the following specific steps: (1) constructing a convolution long-term and short-term memory network, and denoising an original image acquired by a sensor in combination with spatial and temporal information of multiple channels; (2) constructing a residual network, learning a mapping relationship from the denoised original image in the step (1) to an RGB image, and performing demosaicing processing on the image; and (3) constructing a cyclic generative adversarial network, learning conversion from night illumination to daytime illumination, and performing reillumination conversion on the image obtained in the step (2). The method aims at weak light imaging application. In order to solvethe problem that a clear and high-contrast color image is difficult to recover due to an extremely low signal-to-noise ratio of an acquired signal, high-quality, color and real-time imaging of a low-illumination scene is realized by utilizing deep learning in combination with denoising, demosaicing and heavy illumination processing methods, and the method is widely applied to the fields of military affairs, security and protection monitoring, scientific research and the like.

Description

technical field [0001] The invention belongs to the field of computational photography and deep learning, and relates to a low-light color imaging method based on deep learning. Background technique [0002] Low-light imaging has a wide range of applications in surveillance, military, environmental monitoring, scientific research and other fields. In a low-light environment, the optical signal itself is extremely weak, and the current general-purpose camera sensor is not sensitive enough and has high readout noise, so the signal-to-noise ratio of the collected signal is extremely low, and the useful signal is drowned by the noise. [0003] Images recovered using traditional algorithms contain a lot of noise, making it difficult to distinguish the contours and shapes of captured objects, and it is even more difficult to reproduce details and colors. Existing night vision imaging technologies are generally based on image intensifier tube technology and thermal infrared techno...

AI Technical Summary

Problems solved by technology

However, these technologies cannot realize the acquisition of color images (such as low-light night vision devices, thermal infrared imagers), or require active light source irradiation (infrared active light source + color camera imaging), all of which cannot meet the needs of extremely weak light (< Color imaging under 1e-3Lux) conditions

Images