Rotational speed measurement using a low-cost imaging device and image processing algorithms

Abstract

Accurate and reliable measurement of rotational speed is desirable in many industrial processes. A novel method for rotational speed measurement using a low-cost camera and image processing techniques is presented in this paper. Firstly, sequential images are continuously processed using a similarity evaluation method to obtain the periodic similarity level of captured images. Subsequently, the rotational speed is determined from the periodicity of a restructured signal through Chirp-Z transform and parabolic interpolation based auto-correlation, respectively. The measurement principle and system design are presented. The advantages of the proposed measurement system include non-contact measurement, low cost, no markers required and high accuracy. Experimental investigations into the effects of the periodicity detection algorithm, frame rate and image resolution on the accuracy and reliability of the measurement system are conducted on a purpose-built test rig. Experimental results demonstrate that the system with the frame rate of 100 fps yields a measurement error within ±0.6% over a speed range from 100 to 3000 RPM (Revolutions Per Minute). More accurate and reliable speed measurements over a wider speed range are achievable with higher frame rates.