Position and Vector Detection of Blind Spot motion with the Horn-Schunck Optical Flow
Stephen Yu,Mike Wu
DOI: https://doi.org/10.48550/arXiv.1603.07625
2016-03-24
Abstract: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.
Computer Vision and Pattern Recognition
What problem does this paper attempt to address?