Human exploration of the deep sea requires high-quality underwater images to obtain valuable information. However, due to the absorption and scattering of light propagating underwater, underwater images have poor visibility and require further repair or enhancement.
We proposed a novel Retinex-based underwater image enhancement method. The Retinex model is a simulation of the color perception of the human visual system, which can restore colors and adjust lighting to make subjective visual perception more comfortable.
Keywords: Underwater image enhancement, Retinex decomposition, Color correction strategy.

Used a binary detector to differentiate generated fake images from real images

Studied the occurrence and evolution of hematoma and edema in patients with hemorrhagic stroke, and provided suggestions for clinical decision-making

Used transfer learning to train a peach disease detection model using a pre-trained model trained on the ImageNet dataset