Specialists from Smart Engines and scientists from IPPI RAS, Federal Research Center "Informatics and Management" RAS have developed a new image processing method for televisions that allows both people with normal color perception and color blind people to watch content simultaneously.
The method is based on modeling the vision of dichromats, which allows algorithms to identify details that may be difficult for color blind people to perceive. The system adjusts the brightness and enhances the contrast of these elements, while maintaining natural colors, which makes objects recognizable to people with normal vision.
We have been working on this for several years, and our method really provides distinction and preserves the naturalness of the image
In an experiment involving both color blind people and viewers with normal color perception, 95% of trichromats and 90% of color blind people noted that images processed with the new method looked more realistic compared to traditional adaptation methods.
Read more on the topic:
A Neural Network Was Trained in Kazan to Recognize Traffic Light Signals
A Digital Twin of the Eye Developed by Scientists from the Russian Federation