Measurement of the Optical Properties Using Genetic Algorithm Optimized Neural Networks

Back Propagation (BP) neural network was optimized by Genetic Algorithm (GA) to be Genetic Algorithm Optimized Back Propagation (GA-BP) neural network. The data of Monte Carlo simulations was used to train BP neural network and GA-BP neural network. The accuracy of each neural network was investigated and compared. The result showed GA-BP neural network was much steady and more accurate. A laser source and a CCD camera were used to measure the image of diffuse reflectance of bovine muscle, pig muscle and human skin tissue. The optical coefficients of them were determined by the GA-BP neural network. It was found that our results were in good agreement with their reference values.