⚡ 本页包含 AI 生成的分析内容，仅供参考
该论文提出了一种单芯片智能眼动追踪传感器，采样率达5000次/秒。通过红外照明和片上处理直接计算瞳孔中心位置，避免了传统方案中昂贵的外围器件和软件计算，实现了低功耗、高速度的眼动跟踪。
user gazes. Infrared (IR) light is commonly used in an eye tracker because it eliminates the influence of ambient illumination and improves the discrepancy between the pupil and the white area in the eye [1]. With the IR illumination the pupil is the biggest black region in the eye image. Therefore, the address of the point at which the eye gazes can be obtained by finding the center point of the pupil in the eye image. A conventional eye tracker using a normal image sensor needs many peripherals and software computation causing high cost, low speed and high power consumption. As the tracking resolution is increased, the tracking time is increased drastically and is usually limited to about 60S/s because of the speed of the ADC, the data interface and computation [2, 3]. To overcome these
Dongsoo Kim, Jihyun Cho, Seunghyun Lim, Dongmyung Lee, Gunhee Han
Yonsei University, Seoul, Korea An eye tracker is a system that acquires the address of the point at which the