Research On Particle Size Visual Inspection System

Particle size is a process indicator in many important areas now, such as environmentalmonitoring, mineral processing, pharmaceutical and powder metallurgy and so on. Their rawmaterials or products are composed by different particle size, so there is a closely relationshipbetween particle size and the material properties, like pulverized coal’s combustion efficiencyin the combustion process, chemical paint’s adhesion to the wall surface, the particlessuspended in the atmosphere and human health, the size of the pharmaceutical preparationsand the probability of the absorption of the body, industrial materials and its magneticpermeability proportional relationship, the insecticide which applied to farming and its theextent of residues and the effectiveness of drugs and so on, all related to particle size. Inshort, the particle size is the main indicators affecting the material performance, thereforeparticle size test has become an important basic work of industrial materials produced,application, research.In today, the current detection for particle size by sieving method, microscopysedimentation method in most cases, these methods accuracy is low, the measurement iscomplex, the representation is poor. With the rapid development of modern computertechnology, especially combined with the popular image processing and analysis technology isgradual development and improvement, the machine vision detection technology has beenwidely applied in particle size detection system. Particle detection based on image processingtechnology has been great progress compare with other detection methods, there are manyadvantages that can effectively overcome these shortcomings of traditional methods, and at thefinal test, the machine vision detection technology can obtain reasonably accurate results, andimage processing technology in particle size detection system also represents the long-termdevelopment.Firstly, the paper selected diameter of the particles in the sample between50μm to550μm, used high-resolution linear array camera and industrial optical zoom lens to get abetter picture of the particles. secondly, due to the small particle mass and large quantities, itrequires a rapid, efficient method for collecting samples of particle detection, so a high-speedDSP processor is be used. Finally, the particle size of the collected image according to the characteristics makes image processing, and achieves the target of the experiments.Experiments show that, the particle size were detected based on machine vision imageprocessing technology, which has simple and intuitive result, and the image recognitionachieves stable and reliable object.