A 256×256 14k Range Maps/s 3-D Range-Finding Image Sensor Using Row-Parallel Embedded Binary Search Tree and Address Encoder

📋 论文概要

本文提出了一种采用行并行嵌入式二叉搜索树和地址编码器的256×256分辨率3D测距图像传感器，实现了14k range maps/s的高速测距。针对传统飞行时间法速度不足的问题，采用光截面法实现高速高精度3D测距。

💡 主要创新点

• 采用行并行嵌入式二叉搜索树架构，实现像素激活的高速搜索与编码，提升测距帧率。
• 结合地址编码器，有效减少输出数据量，适用于高速3D成像应用。

🏷 关键词

📄 原文摘要

These days, 3-D information technology is being developed rapidly and has been applied to various fields. Moreover, ultra-fast 3-D range-finding makes way for the possibilities of additional applications such as drop tests, observation of a high-speed moving target, and automated controls in robot vision. While 3-D range-finding image sensors using time-of-flight (TOF) don’t achieve high-speed 3-D range finding [1,2], several high-speed and high-accuracy 3-D range-finding image sensors employing the light-section method have been reported [36]. Column-parallel search of a pixel activated by the reflection of a projected sheet beam is repeated the same number of times as the number of rows [3,4]. Row-parallel search of an activated pixel in each row at the same time is suitable for the light-section method when the light-section method uses a sheet beam as a horizontal scanner. Previous work [5,6] employs row-parallel search and

👥 作者与机构

Shingo Mandai, Makoto Ikeda, Kunihiro Asada

University of Tokyo, Tokyo, Japan