4D Computer Vision™
Computer Vision is the use of programming for making selected determinations regarding or utilizing data from digital images.
4D Computer Vision uses measure points to precisely locate selected features from sensor observation-derived representations. Doing so enables 2D video image streams to be used to efficiently perform 3D (depth) and 4D (over time) analysis of selected features from those streams of video images.
4D Computer Vision utilizes the smallest datasets that can be used to reliably provide selected determinations.
4D Computer Vision DOES NOT require the use of Machine Learning, training sets, or simulations.
Developers have the ability to utilize their selections of programming to exercise complete control over the architecture, operations, outcomes, and efficiency of the 4D Computer Vision resources they build.
4D Computer Vision vs. Legacy Computer Vision
Detecting when a driver has fallen asleep highlights the significant operational differences between 4D Computer Vision and Legacy Computer Vision.
4D Computer Vision utilizes a working dataset that includes the X-line location and Y-line location of two precisely located pixels from each sequential image of a driver’s face.
One pixel is located at the lower center of an upper eyelid and the other pixel is located at the upper center of the lower lid of the same eye.
If a distance between those two pixels that indicates closed eyelids remains the same for a certain number of consecutive images or longer, it can be concluded that the driver has fallen asleep.
By comparison, Legacy Computer Vision methodologies must make millions of determinations regarding each video image of a person’s face to recognize that a driver has fallen asleep.
