Rotational Speed Measurement Through Image Similarity Evaluation and Spectral Analysis

Abstract

Accurate and reliable measurement of rotational speed is desirable in a variety of industries. This paper presents a rotational speed measurement system based on a low-cost imaging device with a simple marker on the rotor. Sequential images are pre-processed through denoising, histogram equalization, and circle Hough transform, and then processed by similarity evaluation methods to obtain the similarity level of images. Finally, the rotational speed is obtained through the Chirp-Z transform on the restructured signals. The measurement principle, structure design, and performance assessment of the proposed system are presented. The effects of different influence factors, including frame rate, marker shape and size, an algorithm for image similarity evaluation, illumination conditions, shooting angle and photographic distance, on the performance of the measurement system are quantified and discussed through a series of experimental tests on a laboratory test rig. Experimental results suggest that the system is capable of providing constant rotational speed measurement with a maximum relative error of ±0.6% and a repeatability of less than 0.6% over a speed range from 100 to 900 RPM. Under varying speed conditions the proposed system can achieve valid measurement with a relative error within ±1% over the speed range of 300 to 900 RPM.