ANALYSIS AND SEGMENTATION METHODS OF DIGITAL IMAGES. BORDER OBJECT EXTENSION ALGORITHM

In the production and use of bulk materials, it becomes necessary to control their particle size distribution. The particle size of the bulk material determines important technological characteristics and affects the effectiveness of its application. Classical methods for determining the particle size distribution of agglomerate are laboratory (sieve) analysis and sieving. Along with them, some enterprises use methods based on measuring the energy of primary crushing of the material, mechanical granulometry, and they have also found application of optical methods and implemented systems based on them using samplers. Automation of control of the granulometric composition of bulk materials using the optoelectronic method directly in the process stream will allow you to timely classify the material by size classes and evaluate its strength, which depends on many factors. Recent developments in the field of digital image processing provide the opportunity to get a better idea of the particle size distribution of particles, which is difficult to obtain using mechanical sieve analysis. A method is being developed based on the algorithm of a growing region with strengthening of its edge and in expanding the image as methods of pre- and post-processing. The aim of this study is to develop an image analysis technique to provide information on the full range of particle sizes of an agglomeration charge, starting from small to large sizes.