路侧计算单元 MEC
路侧计算单元（MEC），利用摄像头、雷达等路侧传感设备采集的视频、点云等原始数据，通过内置高性能GPU和AI算法，可有效识别道路中的交通参与者、交通事件、交通流量三项主要信息，其中交通参与者又分为机动车、非机动车、行人三大类，目标均带有速度、航向、位置等关键信息，交通事件包括异常停车、逆行、超速、行人闯入等十余种事件类型，为车路协同系统提供全息感知能力。可广泛应用于城市路口、高速公路、园区、AVP、矿区、测试场等各类车路协同业务场景。
主要特点
支持多源异构传感器接入，具备交通参与者、交通事件等的高精度融合感知
支持面向交管、运输、监管、智能驾驶等业务需求的智能决策和交通基础设施联动
面向路侧边缘计算的规格设计，小尺寸、大算力、低功耗、宽温设计
支持Ethernet/USB/RS232/RS485/CAN/DIO等多种通信接口
支持超低时延计算，通过软硬件一体的高效算法实现毫秒级全息感知
具备自动标定、一体化部署工具、感知效果测评等完善的工具体系
Support multi-source heterogeneous sensor access, with high-precision fusion perception of traffic participants, traffic events, etc.
Support intelligent decision-making and traffic infrastructure linkage for traffic management, transportation, supervision, intelligent driving and other service needs.
Designed for roadside edge computing with the desigh of small size, strong computing capability, low power consumption, wide temperature range.
Support various communication interfaces such as Ethernet/USB/RS232/RS485/CAN/DIO.
Support ultra-low latency computing, achieving millisecond-level holographic perception through efficient algorithms integrated by hardware and software.
Equipped with a complete tool system including automatic calibration, integrated deployment tools, perception effect of evaluation, etc.
