Research On The State Monitoring Of Moving Hot Steel Slabs Based On Images

Speed and temperature are important parameters to achieve automation in the billet heating field. The accurate measurement of temperature is an effective method for improving product quality, conserving energy. The real-time measurement of speed is an effective mean for reducing labor costs, improving control accuracy. Now, manual speed regulation and conventional contact sensor would not meet the demands of industrial measurement. The non-contact measurement using an image sensor is a new technique that integrates electronics and image processing technique, and attracts the attention of the researchers in measurement field on account of its advantages like quick response, non-contact measurement and wide available range. In this paper, based on image processing technology we propose a method to measure the speed and temperature of billet. In this method, we use industrial camera to get image sequence of moving billet.This paper introduces radiation measurement theories and methods. An improved temperature model is presented based on an original radiation measurement model. The proposed method takes exposure time as a variable and deduces the relationship among gray value of image, exposure time and temperature. The original model between gray value and temperature was established by extreme learning machine. By changing exposure time, current temperature model can be predicted using original model. Finally, experiment results by means of simulation on several groups of test data illustrate the efficiency of the proposed temperature model.This paper presents an image-based method for measuring speed. Because of irregularly shaped oxide scale on the surface of billet, a method is proposed to detect the stationary state of billet by binary image and user-defined similarity criterion when billet is at rest. Considering moving billet, a feature point detection algorithm is presented to extract the matched Surf feature points between adjacent images and calculate the speed. Finally, the speed measurement system is established by integrating two algorithms.