A Lossless Compression Method for Event Cameras via Removing Spatial Redundancy

Event cameras are neuromorphic devices with high temporal resolution, high dynamic range, and low power consumption. However, the event stream data generated by these cameras requires a large amount of storage. To address this issue, we propose a lossless compression method. We encode the event data in a specific format by optimizing spatial redundancy and temporal redundancy of the event sequence data. Our method is based on a multi-level dictionary structure. First, the imaging domain is divided into smaller blocks to reduce the size of data type for spatial coordinates. Second, a novel multi-level dictionary is proposed to remove the spatial coordinate redundancy. Finally, a differential array is used to reduce the timestamps redundancy. Our experimental results on both synthesis and real event data show that the proposed method can achieve a higher compression ratio than the state-of-the-art methods. This proposed method can be useful in applications where event data needs to be compressed and stored.