LRGO/MoS₂ heterojunction combined with artificial neural networks for intelligent identification of peach ripeness

Artificial Neural Networks (ANNs) have demonstrated substantial potential in classification and artificial vision due to their self-learning capabilities and optimization efficiency. However, research on their application in multimodal integrated recognition remains limited. This study introduces a novel method to intelligently identify peach ripeness utilizing a dual-functional synaptic sensor composed of LRGO/MoS₂ combined with ANNs. We fabricated a heterostructure of LRGO/MoS₂ exhibiting exceptional optoelectronic and piezoelectric properties to create a multimodal sensor capable of responding to both optical and mechanical stimuli. In terms of photoconductive response, the sensor selectively reacts to different wavelengths of light and adjusts photocurrent according to light intensity and exposure duration. Regarding mechanical pressure response, the LRGO/MoS₂ sensor shows high sensitivity, accurately detecting variations in applied pressure. The sensor's photogenerated current and pressure data are fed into the ANN, enabling precise identification of peaches at varying ripeness levels. After training, the ANN achieved a commendable accuracy of 96% after 500 iterations. This innovative combination of the LRGO/MoS₂ dual-functional synaptic sensor and ANNs heralds a new era for developing low-cost, rapid, and efficient multimodal recognition systems.