Neural Network Based Approach for Automotive Brake Light Parameter Estimation

Abstract

Abstract

Corona segmentation for nighttime brake light detection

“…Considering the recent developments in intelligent transport systems, there are numerous studies regarding the detection of front car lights [15][16][17][18][19][20], brake lights [21][22][23][24][25][26][27][28][29][30], and also many relevant studies [31][32][33][34][35]. As a brief summary of the recent and prominent studies about brake light discrimination which are particularly related to our research, Thammakaroon and Tangamchit [21] proposed an red-green-blue (RGB)-based detection system, which discriminates brake lights using the mean and standard deviation of colour channels, with taking the differences of dark red (tail light) and light red (brake light) signals.…”