← 返回论文列表 📄 下载原文 PDF  ISSCC 2026 · 7.4
ISSCC 2026Session 7 · IMAGE SENSORS AND RANGINGImage Sensors90nm BSI CIS

A 480×320 CMOS LiDAR Sensor with Tapering 1-Step Histogramming TDCs and Sub-Pixel Echo Resolvers

⚡ 本页包含 AI 生成的分析内容,仅供参考

📋 论文概要

本文提出一款480×320 CMOS LiDAR传感器,采用锥形1步直方图TDC和亚像素回波解析器,实现内存高效操作。该传感器在90nm BSI CIS工艺中实现,在100米距离达到3.8cm精度,并在110klx阳光条件下以25fps实现4.1cm准确度,解决了LiDAR在强光下高精度测距和分辨率提升的难题。

💡 主要创新点

核心指标
3.8cm precision at 100m, 4.1cm accuracy under 110klx sunlight at 25fps
工艺节点
90nm BSI CIS
重要性
发表年份
ISSCC 2026

🏷 关键词

LiDAR传感器直方图TDC亚像素回波解析SPADCMOS图像传感器

📄 原文摘要

*Equally Credited Authors (ECAs) 1 Abstract A 480×320 CMOS LiDAR sensor using tapered 1-step hTDCs and SPER for memory-efficient operation is presented. The hTDC accumulates histograms with 1ns resolution over the full range and applies 500ps resolution only at the echoes, while SPER maps echoes to individual SPADs, enabling 4× image resolution. Implemented with 2×2 SPAD macro-pixels and shared AFEs (2.5T/SPAD) in a 90nm BSI CIS process, the sensor achieves 3.8cm precision at 100m and 4.1cm accuracy under 110klx sunlight at 25fps. Light detection and ranging (LiDAR) is indispensable for 3D imaging in autonomous vehicles, robotics, and surveillance systems. For reliable object recognition, LiDAR faces two key challenges: fine time resolution for depth accuracy and high image resolution for detailed spatial mapping. Fine time resolution requires a large number of histogram bins, while high image resolution demands histograms for more pixels. Conventional 1-step

👥 作者与机构

Wonjong Roh*1,2, Canxing Piao*2, Taehoon Jeon1, Junsang Bae1, Jung-Hoon Chun1,2, Seong-Jin Kim3, Jaehyuk Choi1,2

Sungkyunkwan University, Suwon, Korea, 2SolidVue, Seongnam, Korea, 3Sogang University, Seoul, Korea

分类:Image Sensors · 年份:ISSCC 2026