Pedestrian Detection Algorithm in Nighttime Environment Using the Cascade HOG in Lab Space
Jungyu Park, Dae-Seong Kang · Advanced science and technology letters · 2015
In this paper, an algorithm to track night pedestrians in real time is proposed. First, data is converted into a night L * a * b * color space, and then the L area is extracted. This data combined with image subtraction creates pre- processing data. Then, using the Cascade Histogram of Oriented Gradient (HOG) algorithm detects the pedestrian at night.