面向智能图像传感系统的感内计算技术发展趋势分析  

Development trend analysis of in-sensor computing technology for intelligent image sensing systems

作　　者：周正[1] 丛瑛瑛 冯玉林 ZHOU Zheng;CONG Yingying;FENG Yulin(School of Integrated Circuits,Peking University,Beijing 100871,China;Informatization and Industrialization Integration Research Institute,China Academy of Information and Communications Technology,Beijing 100191,China;School of Instrument Science and Opto-Electronics Engineering,Beijing Information Science&Technology University,Beijing 100192,China)

机构地区：[1]北京大学集成电路学院,北京100871 [2]中国信息通信研究院信息化与工业化融合研究所,北京100191 [3]北京信息科技大学仪器科学与光电工程学院,北京100192

出　　处：《信息通信技术与政策》2023年第6期24-29,共6页Information and Communications Technology and Policy

摘　　要：物联网大数据时代,图像传感系统的端侧智能应用大规模部署,催生了传感器向智能化、高集成度和低功耗的趋势发展。感内计算技术利用像素层级的感算、感存算一体化功能,实现了二维空间感知信息的高精度、高并行模数混合运算。感内计算技术结合近感计算和感外计算技术的研究成果,将为面向端侧智能应用的新型图像传感系统提供一系列低成本、多元化解决方案。In the era of Internet of Things and big data,the large-scale deployment in the end side of image sensing systems motivates the sensors to develop towards superior intelligence,high integration,and low power consumption.Using the integrated functions of sensing and storage at the pixel level,in-sensor computing(ISC)can realize the high-precision parallel mixed-signal processing of two-dimensional spatial perceptual information.A series of cost-effective and diverse solutions for advanced image sensing system for intelligent applications in edge can be provided,through combining ISC with near-sensor computing and off-chip post-processing.

关 键 词：物联网 图像传感系统 感内计算技术 

分 类 号：TN492[电子电信—微电子学与固体电子学]