Intelligent all-day vehicle detection based on decision-level fusion using color and thermal sensors
Shih-Che Chien, Feng-Chia Chang, Chiung-Cheng Tsai, Yung-Yao Chen · 2017
Recently, research topics involving vehicle detection become popular for the automotive advanced driver assistance systems (ADAS). Vehicle detection is an important part of unmanned vehicle driving. It is particularly challenging in real-world scenarios due to several uncontrollable factors, such as variances in light, weather, and unpredictable scenes. Most vehicle detection techniques focus on using a visible spectrum color camera to carry out day-time vehicle detection, or using an infrared thermal imaging camera to carry out night-time vehicle detection. However, fewer researches for all-day vehicle detection have been done so far. In this paper, we present a novel all-day vehicle detection method, in which both visible and thermal information are used to detect the vehicles separately, and the final decision is made by decision-level fusion. The experimental results demonstrate that the proposed method can effectively detect vehicles in different environments and achieve robust recognition rates.