Back to results
US 12,597,093 B2Granted
Underwater image enhancement method and image processing system using the same
AssigneeNational Chengchi University
InventorsYan-Tsung Peng, Yen-Rong Chen
Life of this patent
Estimating the term…
Abstract
An underwater image enhancement method and image processing system using the same are provided. The method includes the following steps. An original underwater image is received. An original histogram of the original underwater image is generated. The original histogram is input into a deep learning model to generate an optimized histogram. An optimized underwater image is generated according to the optimized histogram and the original underwater image.
Claims18
An underwater image enhancement method, comprising:
- receiving an original underwater image;
- generating an original histogram of the original underwater image;
- inputting the original histogram into a deep learning model to generate an optimized histogram;
- generating an optimized underwater image according to the optimized histogram and the original underwater image;
- and
- training the deep learning model based on a weighted sum of a plurality of loss functions,
- wherein the loss functions comprise a first loss function, a second loss function, and a third loss function, the first loss function calculates a first loss value according to a reference histogram of a real reference image and the optimized histogram, the second loss function calculates a second loss value according to the target image and the real reference image through an error calculation method, and the third loss function calculates a third loss value according to the target image and the real reference image through another error calculation method.
4 more elements
An image processing system, comprising:
- a storage circuit;
- a processor, coupled to the storage circuit and configured to:
- receive an original underwater image;
- generate an original histogram of the original underwater image;
- input the original histogram into a deep learning model to generate an optimized histogram;
- generate an optimized underwater image according to the optimized histogram and the original underwater image;
- and
- train the deep learning model based on a weighted sum of a plurality of loss functions, wherein the loss functions comprise a first loss function, a second loss function, and a third loss function, the first loss function calculates a first loss value according to a reference histogram of a real reference image and the optimized histogram, the second loss function calculates a second loss value according to the target image and the real reference image through an error calculation method, and the third loss function calculates a third loss value according to the target image and the real reference image through another error calculation method.
5 more elements
Dependencies are inferred from the claim text.