Detection method for pulverized coal injection and particles in the tuyere raceway using image processing

Abstract

Accurate monitoring and characterization of the working condition of pulverized coal injection plays an important role in the operation of the blast furnace. Existing monitoring systems neglect particles in the tuyere raceway and lack robust quantitative analysis. This paper presents the image-based intelligent detection method to extract features of the pulverized coal injection and particles in the tuyere raceway simultaneously and real-timely. Intelligent circle and line detection based on Hough transform is applied to obtain the background template of the raceway image. The merged local segmentation algorithm is used to obtain the regions of pulverized coal and particles. The mass flow of PCI is characterized by the coal feature calculated by the linear weighted method. The support vector machine and the width of the minimum enclosing rectangle are finally employed to obtain the size distribution of particles and determine what these particles are mainly composed of. Massive raceway images captured from 15 raceways of a 2 500 m3 blast furnace were used to evaluate the detection method. The results demonstrate that the method is effective to reflect the working condition of PCI and can obtain accurate size distribution of particles in the raceway image.