Convolutional decoding of thermographic images to locate and quantify honey adulterations

摘要: In this research, 56 samples of pure honey have been mixed with different concentrations rice syrup simulating a set adulterated samples. A thermographic camera was used to extract data regarding the thermal development honey. The resulting infrared images were processed via convolutional neural networks (CNNs), subset algorithms within deep learning. CNNs trained and optimized using these detect commonly elusive in as low 1% weight, well quantify it. Finally, model successfully validated which initially isolated from training database. result an algorithm capable identifying floral origins quantifying accuracies 95% 93%, respectively. Therefore, complemented analysis shown be compelling tool for control food quality, thanks traits such high sensitivity, speed, being independent highly specialized personnel.