Event-based improved FAST corner feature detection algorithm

In the era of Industry 4.0, the smart driving and unmanned driving have become hot technologies. According to the principle of biological optic nerve, a new sensor--event camera is designed. The camera has the Characteristics of high dynamics, low data redundancy and low delay. And it’s suitable for real-time target detection and tracking. In this paper, an improved event-based FAST corner feature detection algorithm is pro-posed. With the improvement of the FAST corner detection algorithm, an adaptive FAST corner detection algorithm is proposed to improve the anti-noise performance and robustness of the algorithm. And in the case of high speed rotation, the target can be well tracked. Finally, experiments based on simple and complex, translation and rotation scenes verify that the improved event-based FAST corner feature detection algorithm has better robustness and real-time performance.