Dynamic non-line-of-sight imaging system based on the optimization of point spread functions.

Chengquan Pei,Jiamin Wu,Anke Zhang,Lu Fang,Yue Deng,Feihu Xu,Hui Qiao,David U-Lei Li,Qionghai Dai

doi:10.1364/oe.439372

Abstract

Non-line-of-sight (NLOS) imaging reveals hidden objects reflected from diffusing surfaces or behind scattering media. NLOS reconstruction is usually achieved by computational deconvolution of time-resolved transient data from a scanning single-photon avalanche diode (SPAD) detection system. However, using such a system requires a lengthy acquisition, impossible for capturing dynamic NLOS scenes. We propose to use a novel SPAD array and an optimization-based computational method to achieve NLOS reconstruction of 20 frames per second (fps). The imaging system's high efficiency drastically reduces the acquisition time for each frame. The forward projection optimization method robustly reconstructs NLOS scenes from low SNR data collected by the SPAD array. Experiments were conducted over a wide range of dynamic scenes in comparison with confocal and phase-field methods. Under the same exposure time, the proposed algorithm shows superior performances among state-of-the-art methods. To better analyze and validate our system, we also used simulated scenes to validate the advantages through quantitative benchmarks such as PSNR, SSIM and total variation analysis. Our system is anticipated to have the potential to achieve video-rate NLOS imaging.

Highlights

Imaging hidden objects from a camera’s view is an emerging research topic with a broad range of applications, including robotics vision, remote sensing, medical diagnosis and autonomous vehicles
A non-line-of-sight (NLOS) imaging system sequentially performs the following steps: 1) the pulsed-laser shines on the relay wall, 2) the scattered light is bounced on the hidden object and diffusely reflected to the relay wall, and 3) a transient imaging device records the scattering response from the relay wall
Many methods have been proposed to increase the spatial resolution of the NLOS reconstruction, such as real-time transient imaging for amplitude modulated continuous wave LIDAR applications [50], analysis of missing features based on time-resolved NLOS measurements [51], convolutional approximations to incorporate priors into filtered back projection (FBP) [52], occlusion-aided NLOS imaging using single-photon avalanche diode (SPAD) [21,22], Bayesian statistics reconstruction for accounting for random errors [53], temporal focusing methods [31], compressed methods with different acquisition schemes [54]

Summary

Introduction

Imaging hidden objects from a camera’s view is an emerging research topic with a broad range of applications, including robotics vision, remote sensing, medical diagnosis and autonomous vehicles. Several image reconstruction methods, such as occlusion aware lightinverse methods [21,22,23,24,25,26], deep-learning methods [27], heuristic solutions [28,29,30,31], linear inverse methods [32,33,34], data driving algorithms [35,36,37], back-projection methods [8,9,30,38,39,40,41], and deconvolution methods [2,5,11], were adopted to solve NLOS imaging problems [42]. Such prior information is hard to acquire when solving realistic problems

Objectives

Methods

Discussion

Conclusion