Servo System Control Using Gestures

This paper is devoted to designing a Convolutional Neural Network (CNN) to control a servo system by recognizing gestures. The design will be in Matlab and a Programmable logic controller (PLC). In the first step, the CNN is designed and verified in Matlab. Then the structure of the CNN is exported into PLC, where a general processing algorithm using the imported structure takes place. The trained CNN recognizes gestures and controls a one axis servo system using Motion Function Blocks (MFB). The paper will demonstrate the results with experiments.