Recognition of Suspension Liquid Based on Speckle Patterns Using Deep Learning

We presented a machine learning-based method to recognize suspension by distinguishing dispersoid-dependent speckle patterns using convolutional neural network. The dispersoid size and concentration-related transmissive were recorded lensless camera when coherent He-Ne laser irradiated the suspension. Firstly we realized recognition of polystyrene microspheres-dispersed suspensions with different particle sizes several common including protein powder milk similar concentration. Further from three food unknown concentration was achieved high accuracy 99%. experimental results confirm that this non-contact could find applications in measurement classification liquid containing micrometers-sized particles.