Moisture Content Determination in Walnuts by Using a Sound Detector: Preliminary Results

This study explores a sound processing technique for non-destructive moisture content determination in walnuts. Moreover, there is a need for technology that accurately measures moisture content, with the possibility to be portable, inexpensive, and adaptable to various types of food. Current low-cost methods lack precision, while accurate methods are costly and impractical for small-scale use. The method uses a soundproof box with a microphone connected to a laptop to record the sound of walnut samples dropped through a hole in the top of the box. Twenty walnut samples, with moisture content ranging from 23.5% to 45.9% and weights from 17.4 to 31.6 grams, were dried in a convection oven at 40°C for 26 hours to reach 8% (MC). Then, the 20 samples were additionally dried until fully desiccated at 103°C for 16 hours. The sound recordings were analysed using Audacity software, Sampling (t1, Hzl, Db) to determine moisture content. The results showed a positive correlation between measured and predicted moisture content, indicating the method's potential as a low-cost, real-time moisture determination tool, offering significant time and cost advantages over traditional methods.