Meta-DCP: simple and efficient single-frame image enhancement algorithm for infrared metalens images

Infrared metalenses are attracting intensive interest, as alternatives to traditional infrared optical systems comprising multiple lenses and diffractive elements. However, the images captured by infrared metalenses are often blurry due to aberrations and noise. To improve the performance of these images, we propose Meta-DCP, a concise and single-frame infrared metalens image enhancement algorithm based on the dark channel prior (DCP). We introduce an image degradation model to simulate the degradation process of infrared metalens images effectively. Using this degradation model with the DCP, we can directly recover a deblurred image. Meanwhile, we present a brightness enhancement method with integrating the original image pattern to compensate for brightness loss in the deblurred image. Experimental results on a dataset captured with our infrared metalens demonstrate the effectiveness and superiority of our Meta-DCP.