累加平均提升点扩散函数信噪比实现透过散射介质三维成像

Three-Dimensional Imaging Through Scattering Media by Improving Signal-to-Noise Ratio of Point Spread Function Using Cumulative Average

光学三维成像在医学诊断、空间遥感、人工智能等领域具有广泛且重要的应用。然而,在散射环境中,入射光的波前相位会发生严重畸变,导致传统光学三维成像性能劣化,甚至失效。目前,研究透过散射介质成像大多使用的是昂贵的科研相机。为了降低成本,本团队提出了一种采用普通工业相机进行数据采集的透过散射介质的三维成像方法。针对普通工业相机采集的低信噪比图像,本文对代表成像系统点扩散函数的点光源散斑进行多次累加平均,获得了可与科研相机相媲美的高信噪比点扩散函数;然后,将该点扩散函数与单次曝光获得的结构光散斑进行反卷积运算,得到受物体面形调制的变形条纹;最后,基于傅里叶变换轮廓术恢复物体的三维信息。实验结果表明,该方法可以获得与科研相机相媲美的成像效果,为透过散射介质的三维成像提供了一种可行的低成本方案。

Optical three-dimensional(3D)imaging has been widely applied in fields such as medical diagnosis,remote sensing,and artificial intelligence.However,in the scattering environment,the wavefront phase of incident light is severely disturbed,which results in deteriorated performance or even failure of traditional optical 3D imaging.Moreover,the studies of imaging through scattering media mainly use expensive scientific cameras for data acquisition.Therefore,we propose a low-cost 3D imaging method through scattering media based on a common industrial camera for data acquisition.To improve the signal-to-noise ratio(SNR)of the images collected by the common industrial camera,we conduct multiple rounds of accumulative averages on the point light source speckles representing point spread function of the proposed imaging system to obtain a point spread function with an SNR as high as that by scientific cameras.Then,the deconvolution of the point spread function with the structured light speckle obtained by a single exposure is carried out to produce the deformed fringe pattern modulated by the object shape.Finally,the 3D information of the object is recovered by Fourier transform profilometry.The experimental results indicate that the proposed method can achieve an imaging effect comparable to that by scientific cameras,which provides a feasible and low-cost solution for 3D imaging through scattering media.