Research On Lensless Coded-Mask Imaging Technology For Visible Light

The Internet of things(IoT)is a highly integrated and comprehensive application of the new generation of information technology,which is of great significance for in-dustrial transformation and economic growth.Among all these sensors using in IoT,the image sensors are so important that can be called "eyes of all things".If we want to install cameras anywhere,especially in some space-constrained places,the camera must be thin,compact,cheap and the like.However,because of the quantity and thickness of lenses,the focusing distance and flange distance,the thickness of traditional imaging system can't be further reduced.Lensless imaging technology not only can reduce the thickness of the system effectively,but also can reduce the volume and weight.It can be applied to many applications,such as embedded imaging,wearable devices and so on.This thesis mainly study the coded-mask computational imaging which can be used in extended scenes at visible wavelengths.The separable coded mask is directly placed atop of the image sensor instead of the lens.The coded mask encodes the scene light information and then uses algorithm to reconstruct the scene.The main research work is as follows:1.Study the coded mask imaging for extended scenes at visible light.Analy-sis coded mask theory and study relevant optically encoded imaging models and light propagation models for different coded apertures.2.Study separable coded mask.Use m-sequences to constructed separable coded matrix,and conduct theoretical analysis and simulation of diffraction effects of coded mask imaging to determine the relevant physical parameters of the imaging system.According to the mathematical model of separable coded mask imaging,obtaining the transformation matrix of the imaging system by image calibration method which can greatly reduce the difficulty of solving the model and improve the calculatinig speed.3.Study image reconstruction algorithm of coded-mask imaging.When the sys-tem transformation matrix is known,the problem can transformed into an ill-conditioned inverse problem.We use SVD-Tikhonov and TV regularization algorithms to recon-struct the original scene image.4.Build experimental platform.Build lensless imaging prototype at visible wave-lengths.We successfully reconstruct the real scene and obtain 512 x 512 gray images,verify the feasibility of imaging.