Measurement of materials refractive indexs based on laser self-mixing interferometry and convolutional neural network

The refractive index of a material is one of the most important optical parameters. In this paper, we propose the method of Self-Mixing Interferometry (SMI) to measure the refractive index of materials. SMI is superior to other laser interferometry methods because of its characteristics of simplicity and compactness. However, SMI signals are not easy to be analyzed due to the low signal-to-noise ratio and the loss of phase information. Based on the advantages of Convolutional Neural Network (CNN), in this work, we propose a scheme to reconstruct the refractive index of materials from SMI signals based on CNN. With the injection current to the laser being driven by a sawtooth wave, we first obtain different SMI signals by letting the light passing through materials with different refractive indexes under the condition of known material thickness, and then train CNN with SMI signals. The trained network is then used to estimate the refractive indexes of materials. The results show that the method is noise-proof and has high adaptability to the measurement under different conditions.