Real-Time Analog Pixel-to-Pixel Dynamic Frame Differencing with Memristive Sensing Circuits

Abstract

In this paper, we propose an analog pixel differencing circuit for differentiating pixels between frames directly from CMOS pixels. The analog information processing at sensor is a topic of growing appeal to develop edge AI devices. The proposed circuit is integrated into a pixel-parallel and pixel-column architectures. The proposed system is design using TSMC 180nm CMOS technology. The power dissipation of the proposed circuit is 96.64mW , and on-chip ares is 531.66µm 2 . The architectures are tested for moving object detection application.

Building a Real-Time 2D Lidar Using Deep Learning

“…Working at the stage of sensor is also used for depth prediction: OPA (offset pixel aperture) also called DP (duelpixel) cameras [28]; these cameras have spilt green pixels to two halves that could be used for depth estimation combining with machine learning for enhanced estimations (with a single camera [29] and with a stereo [30]). Analog pixel-to-pixel conversion is used for object and motion detection [31]. It works on pixels before they get digitalized.…”