Improving processing time for visual measurements of displacements of IPMC actuators using CUDA

A Graphics Processing Unit (GPU) based measuring system which is used for processing images from a camera to provide information about displacement is presented in this work. The proposed approach has been developed for measuring small movements of micro robotic systems based on synthetic IPMC (Ionic Polymer Metal Composites) materials using CUDA (Compute Unified Device Architecture) technology. The technique is based on processing the data from an analysis of subsequent images of the moving item using Graphics Processing Units in real time. The processing computation can get a more than 3× speedup and the software implements a graphic display of the movement in relation with time. The main improvement is the implementation of real time system and the reduced processing time between two consecutively retrieved frames with satisfactory accuracy.