Tracking System for a Coal Mine Drilling Robot for Low-Illumination Environments

In recent years, discriminative correlation filters (DCF) based trackers have been widely used in mobile robots due to their efficiency. However, underground coal mines are typically a low illumination environment, and tracking this environment is challenging problem that has not adequately addressed the literature. Thus, paper proposes Low-illumination Long-term Correlation Tracker (LLCT) designs visual system for mine drilling robots. A low-illumination framework combining image enhancement strategies long-time proposed. long-term memory filter tracker with an interval update strategy utilized. addition, local area detection method proposed prevent failure of algorithm over-exposure. convenient boost Extensive experiments on popular object benchmark datasets demonstrate significantly outperforms baseline trackers, achieving high real-time performance. The tracker’s performance verified robot mine. results field experiment novel better than state-of-the-art environments.