鬼影成像
计算机科学
帧速率
光学相关器
人工智能
计算复杂性理论
像素
计算机视觉
二进制数
光学成像
加速
帧(网络)
光学
物理
算法
傅里叶变换
数学
电信
并行计算
算术
量子力学
作者
Ayano Inoue,Ren Usami,Keisuke Saito,Yasunobu Honda,Kanami Ikeda,Eriko Watanabe
标识
DOI:10.7567/1347-4065/ab2f6b
摘要
Computational ghost imaging (CGI) has received increasing attention as a single pixel imaging technique. However, the application of the CGI technique to moving objects requires further investigation because it requires a large number of patterns at high speeds. In this study, we propose and experimentally demonstrate optical correlator-based CGI as a step towards high-speed CGI. Our optical correlator plays the role of a spatial light modulator and correlates a target object to known patterns of general computational imaging. We experimentally verified that the RMS error of the reconstructed images is similar to that of the numerical calculations. If one reconstructed image is acquired using 1000 random binary patterns, the imaging frame rate of 133.7 fps will be achieved. This provides high-speed CGI applicable to moving objects and other phenomena.
科研通智能强力驱动
Strongly Powered by AbleSci AI