A Measurement Method Of Pellets Diameter Based On Wavelet Transform

The fired pellet production is a main part of the modern ironmaking processing. The particle-size of the pellets and its distribution is also an important index in the quality control of the iron-ore pellet production. The traditional pellets diameter measurement means have many deficiencies. With the development of the computer and artificial intelligence technology, the computer vision technology which plays an important role in the image processing provides a new way for the particle size examination.This paper first makes a brief summary of the phylogeny of the wavelet, the wavelet theory and application of the wavelet to the image processing, then does a brief introduction about the on-line measuring the pellet diameter computer system. The system consists of three parts, i.e. pellet image pretreatment, pellet image segmentation and pellet particle-size measurement. Firstly, based on the analysis of the features of the pellet image, an image denoising method using wavelet packet transform and median filter is presented, which meets the demands of the following pellet image processing. Then in view of the shortage of the traditional edge detection measures, a method of edge detection based on the wavelet transform is presented, which makes good results and proves the superiority of the wavelet transform in the edge detection. Finally, the particle size is measured based on the image segmentation and circularity theory. The results of the experiments show that the method is feasible and the particle-size distribution obtained from the presented method is close to the real measured results. This system can be applied into site application.