Research On The Optimization Of Surface Plasmon Resonance Imaging By Noise Analysis

Surface plasmon resonance imaging(SPRi)is a kind of novel optical imaging technique based on the surface plasmon resonance(SPR)effect supported by the nano structures of noble metals.It features a character of label-free,high sensitivity and high throughput imaging ability,and because of the novel detection method and outstanding performance,SPRi enjoys both great importance in theoretical research and extensive application prospect in multiple fields.When SPR takes place,the electromagnetic field in the vicinity of the metal structure will be drastically enhanced,while sharp dip will occur on the reflected spectrum,which is extremely sensitive to the variation of Refractive Index(RI)on the metal-dialectic interface.SPRi thus measure the change of sample RI by measuring the change of parameters relative to the reflected spectrum.Comparing with those based on intensity and phase interrogation,SPRi based on wavelength and angular interrogation(all belongs to spectral SPRi)feature better linearity and wider dynamical range,and thus are widely used.However,being affected by the combination of the light source,the optical components,the Surface Plasmon Polariton(SPP)excitation module and the CCD sensor,the imaging system suffers from a high noise level.This considerable noise will inevitably deteriorate the image quality and system stability,which will consequently degrade the RI resolution of the imaging system.Unfortunately,available methods for the SPRi optimization realize noise reduction by self-reference technique mostly.These methods either improve the RI resolution at the cost of spatial resolution,or partially reduce the noise level while perplexing the overall noise distribution,which poses obstacles for further system noise analysis and post-processing.Furthermore,although time averaging may reduce measurement uncertainty,it also delays the measuring time,and indirectly raises the dependency of imaging system on the measuring environment.This project aims to systematically analyze the imaging process of SPRi from perspective of the constitution of system signal,by studying the signal change through the transferring from light source to SPP excitation module and finally to the CCD sensor,and then analyzing the individual influence of each steps on the output image signal,based on which,we may conduct system optimization from different angles.Therefore,we thoroughly analyzed the source,characters,influence and reducing methods of different noise components of the imaging system,and then constructed a complete signal model together with a state-space representation.With the help of such a state-space model,we may introduce multiple signal processing techniques to conduct filtering and denoising for the images,to extract pure information of the reflected spectrum,and finally to improving the RI resolution of the system.On the other hand,by analysis of the noise components,we may obtain the approximate probability distribution of the overall system noise,and compute the distribution parameter.By such a statistical distribution,we may simulate the noise of the experimental system,and then simulate the imaging process of it.To address the influence of light source and CCD,we constructed an experimental aided simulation model based on the thought in question.This model allows us to theoretically study the trend of system performance to the change of measuring quantity,and further to predict the optimal system parameter to improve the system performance,by which we will realize optimization for the SPRi.In order to achieve this goal,we explained the principal and procedure of the optimization in detail,and designed a spectral SPRi based on Kretschmann configuration as a verification of the proposed thought.By comparison between experimental and simulation results,we might witness the effectiveness and feasibility of the method,we as well discussed the constraining factors for the optimal parameter of the SPRi system.