Eye Tracking Technology of SEMI to reduce the power required for the virtual reality

Among the critical issues facing the reality systems is certainly the amount of data that needs to be generated in real time to render the correct image to each eye. SMI offers a technology that should help reduce the load on our systems work. The technology part of the assumption that our vision is high resolution only in a central portion of what we see, corresponding to about 5 ° field of view: more devices in the parts of the image quality is much lower.

According to SMI is therefore useless to render the highest quality throughout the image that is presented to our eyes: on the contrary tracing the user's gaze and optimizing the resolution only at the point where he is actually looking at it is possible to save a large amount of calculations system . In practice, the image is decomposed into three circles: one relating to the image portion in which serves the maximum resolution (the smallest), one more external with quality to 60% and an even external to the lowest quality. Moving the circle to the highest quality where the eye rests, our brain should not perceive the difference between this stream to the lower resolution and that rendered 100%.

We tried the technology live and actually it was difficult to discriminate the difference between when the technology was on (saving computing power) and even though it was not active and therefore with the greatest burden on system resources. At the outset the technology seems to work well, it will be interesting to see if some big manufacturer decides to adopt it.