Fast Single-pixel Imaging Based On Frequency Division Multiplexing

Single-pixel imaging can surpass the limit of sampling theorem utilizing sparse constant as implicated by compressive sensing and becomes the hot topic in recent research.However,single-pixels imaging cannot achieve fast speed,good quality and a high resolution,which prevent the widely application.Based on the foundation of conventional Fourier spectrum acquisition means,a new method based on frequency division multiplexing has been proposed.By this method,we can achieve the twice or more speed of acquisition in period of imaging.We also build a fast singlepixel imaging system corresponding to our method to demonstrate the method we proposed.Thesis work mainly includes the following two aspects:1.A fast imaging method based on frequency division multiplexing has been proposed,by this method,we reduce the number of illumination patterns,modulated by SLM,used for acquire the spectrum of object,without change the speed of SLM,we achieve the twice speed than before.2.Based on the method we proposed,we build an optical single-pixel imaging system and do several experiments on it.According to the result of experiment,we achieve 1.85 times and 3.2 times faster than conventional method.The result of experiment indicted that the method we proposed can increase the speed of imaging and work well.