Online Measurement for the Thermal Geometric Parameters of Hot Large Forgings

Abstract

t is difficult to obtain high measurement accuracy in online detecting the thermal geometric parameters of hot large forgings, based on machine vision method. In this paper, firstly, a spectrum selective image acquisition method for high-temperature components is proposed, which can successfully reduced the effect of high temperature on image quality. Then, a measurement method based on the coding feature lines for geometric parameters of hot large forgings is employed, which can effectively improve the extraction efficiency of geometry characteristic information of high-temperature components. Experiments on the forging workshop show that measurement accuracy of diameters for cylindrical forging reaches 0.2% and the measurement time is 2.25 seconds; measurement accuracy of lengths for cuboid forging reaches 0.48% and the measurement time is 5.9 seconds. The experimental results have verified the feasibility and high efficiency of the proposed method for measuring the geometric parameters of hot large forgings.