TY - GEN
T1 - Active Event-Based Static Perception Through Bio-Inspired Blinking Under Extreme Lighting Conditions
AU - Gao, Boyang
AU - Cui, Yangjie
AU - Wang, Ziyu
AU - Tu, Zhan
AU - Dong, Xin
N1 - Publisher Copyright:
© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2026.
PY - 2026
Y1 - 2026
N2 - Event cameras offer exceptionally high temporal resolution and dynamic range, enabling robust capture of scene changes even under low-light conditions. However, their inability to generate events in static scenes fundamentally limits their applicability to tasks such as SLAM, object tracking, and recognition. We propose Robo-Eyelids, a bio-inspired system employing a mechanical shutter to mimic artificial “blinking”. A Random Temporal-Spatio Decay (RTSD) algorithm detects scene stationarity to trigger blinking, producing full-frame event streams, while a formula-based intensity reconstruction generates corresponding intensity images. Furthermore, a Peripheral Trilinear Event Double Integral (Peripheral-TEDI) model fuses APS frames with event data to dynamically select the optimal frame. In static environments, Robo-Eyelids achieves continuous high-quality intensity images at 15 fps under strong light and 8 fps under low light, mitigating the limitations of event cameras in static perception and demonstrating robust performance in downstream tasks such as optical flow tracking and object detection.
AB - Event cameras offer exceptionally high temporal resolution and dynamic range, enabling robust capture of scene changes even under low-light conditions. However, their inability to generate events in static scenes fundamentally limits their applicability to tasks such as SLAM, object tracking, and recognition. We propose Robo-Eyelids, a bio-inspired system employing a mechanical shutter to mimic artificial “blinking”. A Random Temporal-Spatio Decay (RTSD) algorithm detects scene stationarity to trigger blinking, producing full-frame event streams, while a formula-based intensity reconstruction generates corresponding intensity images. Furthermore, a Peripheral Trilinear Event Double Integral (Peripheral-TEDI) model fuses APS frames with event data to dynamically select the optimal frame. In static environments, Robo-Eyelids achieves continuous high-quality intensity images at 15 fps under strong light and 8 fps under low light, mitigating the limitations of event cameras in static perception and demonstrating robust performance in downstream tasks such as optical flow tracking and object detection.
KW - Bio-Inspired
KW - Event Camera
KW - Extreme Light
KW - Static Scene Perception
UR - https://www.scopus.com/pages/publications/105042872356
U2 - 10.1007/978-981-95-8435-2_50
DO - 10.1007/978-981-95-8435-2_50
M3 - 会议稿件
AN - SCOPUS:105042872356
SN - 9789819584345
T3 - Lecture Notes in Electrical Engineering
SP - 611
EP - 625
BT - Proceedings of 2025 9th Chinese Conference on Swarm Intelligence and Cooperative Control - Swarm Control Technologies
A2 - Wang, Qing
A2 - Dong, Xiwang
A2 - Song, Peng
PB - Springer Science and Business Media Deutschland GmbH
T2 - 9th Chinese Conference on Swarm Intelligence and Cooperative Control, CCSICC 2025
Y2 - 31 October 2025 through 3 November 2025
ER -