Research On Optical Properties Simulation And Parameters Inversion Method Of Turbid Medium

The transport of photons in a turbid medium involves the absorption and scattering of photons by medium,which are expressed by the absorption coefficient and reduced scattering coefficient respectively.The study of the transmission law of photons in turbid media and the measurement of optical properties(absorption coefficient,reduction scattering coefficient)of the medium help people to understand the correlation between the physical and chemical properties of turbid medium and its optical properties,so as to provide basic support for the research and development of optical perception technologies such as non-destructive testing of food and agricultural products,and non-invasive medical diagnosis and treatment.At present,simulating the transport process of photons in complex structural turbid media(such as layered tissues,tissues containing heterogeneous)still faces challenges,while the non-destructive measurement accuracy of optical properties is still low.The solution of above problems will promote the application of optical perception technology in food,agricultural product quality inspection,biomedicine,and other fields,and help with the development and optimization of various optical inspection equipment.In this paper,the shortcomings of optical properties inversion and photon transmission simulation of heterogeneous turbid media are analyzed.A sequence-based inversion regression model and improved Monte Carlo simulation method are proposed in combination with Monte Carlo simulation algorithm,which realize the accurate estimation of optical properties and photon transport simulation of heterogeneous turbid media,respectively,and evaluate the detection performance of actual optical instruments.The research contents are as follows:1.A non-destructive measurement(estimation)method of optical properties based on spatially resolved diffuse reflectance technology is proposed,integrating long and short term memory(LSTM)and attention mechanism.In this method,the surface diffuse reflection light intensity curve of turbid medium with known optical properties is obtained through Monte Carlo simulation,which is used to construct the training sample set.The sliding window technology is used to divide the surface diffuse reflection light intensity curve into multiple continuous and partially overlapping sub-intervals as the time-like input of the LSTM module.The output of LSTM(estimation of optical properties)is fused by an attention mechanism.The results of estimating the optical properties of 36 groups of tissue simulation liquid collected by hyperspectral imaging system show that compared with the traditional mechanism model and machine learning model,the proposed LSTM-attention model can improve the estimation accuracy of optical properties(the average relative errors of absorption and reduction scattering coefficients were 14.89% and 9.76%,respectively).2.An improved Monte Carlo(IMC)simulation method is studied for accurate and fast simulation of photon transport in heterogeneous(multilayer,tissues containing heterogeneous)turbid media.Py Qt5 is used to build simulation software.IMC first constructs the medium structure with Cartesian coordinate system to increase the accuracy of the description of the morphology of heterogeneities,introduces the absorption fraction to describe the matching degree of the refractive index of the tissue boundary,and realizes the simulation of photon transmission in a heterogeneous medium by analyzing the position,direction and normal vector of photon contact curved boundary or regular boundary.The simulation results based on four internal heterogeneities showed that the time consumption of IMC was reduced by about 53% when it achieved accuracy consistent with the standard Mesh-based Monte Carlo simulation method(MMC)(the normalized root-mean-square errors of luminous flux were0.51%,0.96%,0.72% and 1.79%,respectively,and the difference in absorption fraction was less than 4%).3.Aiming at the practical application needs of non-destructive detection of internal defects in typical turbid medium such as biological tissues,the influence of parameter setting and defect size on the detection performance of the line scan imaging system is evaluated by IMC method and optical density algorithm,and the parameter optimization and application scope of the imaging system are guided.The results show that the imaging system can achieve better defect detection effect under the condition of light source incidence angle of15° and light source-detector distance of 1mm.For ellipsoid defects at three scales: large(a =2 mm,b = 3 mm,c = 1 mm),medium(a = 2 mm,b = 2 mm,c = 1 mm),and small(a = 2 mm,b = 1.5 mm,c = 1 mm),the defect depth detection limits of the system were 3.5 mm,3 mm,and 2.7 mm,respectively.