Adaptive Slicing Method of the Spatiotemporal Event Stream Obtained from a Dynamic Vision Sensor

• DOI: 10.3390/s22072614
• Publication Date: 2022-03-30T01:45:51Z
• Authors: Yisa Zhang; Yuchen Zhao; Hengyi Lv; Yang Feng; Hailong Liu; Chengshan Han
• Keywords: Motion blur

The dynamic vision sensor (DVS) measures asynchronously change of brightness per pixel, then outputs an asynchronous and discrete stream spatiotemporal event information that encodes the time, location, sign changes. has outstanding properties compared to sensors traditional cameras, with very high range, temporal resolution, low power consumption, does not suffer from motion blur. Hence, have considerable potential for computer in scenarios are challenging cameras. However, visualization is incompatible existing image processing algorithms. In order solve this problem, paper proposes a new adaptive slicing method stream. resulting slices contain complete object information, no can be processed either event-based algorithms or by constructing into virtual frames them We tested our using public as well own data sets. difference between entropy slice ideal less than 1%.