Convolutional Neural Network to Detect the Optimal Water Content of Cassava Chips During the Drying Process

Cassava chips are used as raw materials to manufacture modified cassava flour. To produce high-quality flour, a drying process for is required optimal water content in the range of 15-18% wb. This study aims detect during hybrid hot-air tray dryer with computer vision using convolutional neural network. Three categories chips' wet (water 55-70% wb), semi-dry (20-40% and dry (15-18% wb). In this study, performance four types pre-trained network, i.e., AlexNet, GoogLeNet, ResNet-50, SqueezeNet, were compared by different optimizers (SGDm, Adam, RMSProp) learning rate values, 0.00005 0.0001, resulting 24 experimental design. The results showed 12 network models perfect validation accuracy. AlexNet SGDm optimizer was determined best model because its stable training iteration that experienced no fluctuations, accuracy, specifically 100%, well testing accuracy fastest time, notably 32 minutes. will later be develop rapid, real-time, accurate maintain chip products content.