Chinese Herbal Medicine Recognition Using a VCSEL-Based Time-Delay Reservoir Computing System

We numerically investigate the performance of Chinese herbal medicine (CHM) recognition using a time delay reservoir computing (TDRC) system consist vertical-cavity surface-emitting laser (VCSEL) subjected to optical feedback. A database 960 images from 12 CHM categories is used evaluate performance, most which contain many samples and have different backgrounds. The image feature vector masked optically injected into VCSEL, VCSEL output in two polarization modes are sampled as virtual node states for readout. After testing TDRC under injecting same information we further test parallel processing by half each modes. By concatenating collected readout, speed doubled. Meanwhile, influence key parameters on investigated relationship between idle state also examined. results show that, orthogonal feedback, achieves minimum error 1.7% at rate 2×10 9 per second.