The proposed method uses live image footage which, based on calculations of pixel motion, decides whether or not an object is in the blind-spot. If found, the driver is notified by a sensory light or noise built into the vehicle's CPU. The new technology incorporates optical vectors and flow fields rather than expensive radar-waves, creating cheaper detection systems that retain the needed accuracy while adapting to the current processor speeds.
@article{arxiv.1603.07625,
title = {Position and Vector Detection of Blind Spot motion with the Horn-Schunck Optical Flow},
author = {Stephen Yu and Mike Wu},
journal= {arXiv preprint arXiv:1603.07625},
year = {2016}
}