Online High-Precision Vision Measurement Method for Large-Size Gear Parameters

Abstract

Gear is not only the power transmission system parts of vehicles, machine tools, motorcycles and aircraft, but also the key parts of ship propeller, mining machinery, steel rolling equipment and other machines. Parameter measurement is an important step to ensure gear quality in production, but traditional three coordinate measuring machine(CMM) cannot meet the requirement of real-time online measurement. In this paper, a measuring method of large-size spur gear with high resolution line-scan camera and high precision rotary table is proposed, which can make up the deficiency of small field of view and low precision of the area CCD camera and realize the online precise measurement of large-size gear. The core of this method includes three steps. Firstly, we present a calibration method of line-scan camera to rotary table. Secondly, we present an optimized robust regression algorithm for circle fitting. Finally, we present an algorithm to extract the subpixel tooth profile of noisy gear image. This method can measure the inner diameter, addendum circle diameter, root diameter, roundness, concentricity, circular run-out, common normal, thickness, flatness, chamfer and burr of gear. The experimental results show that the measurement period of this method is 7 seconds, the repetition accuracy of main parameters is 0.002mm, and the measurement accuracy is 0.02mm. Both precision and speed meet the requirements of online measurement in large-size spur gear production.