Research On The Technology Of The Fine Crushing Ore Particle Size Distribution On-line Detection Based On Machine Vision

The information of particle size is the important index in the process of mineral separation,the particle size distribution directly reflects the quality of the comminute process,it is very important to adjust the parameters of mineral processing equipment.At present,most of classical method,for example,sedimentation and fractionation technique are off-linemeasurement.There are some disadvantages in these methods,they take many steps and cost much time,and particle size information doesn't feedback in time.To solve these problems,this paper uses the machine vision technology to collect the ore image on the belt conveyor,detects image of particle size on real-time,draws the particle size distribution curve to analyze,provides information to the optimal control parameters for equipment in the process of mineral separation,and reduces energy consumption,so as to improve the work efficiency of equipment.The main achievements of this thesis are as follows:(1)After analyzing the characteristics of the ore images,which are the noises,low contrast etc.Taking bilateral filtering method and the median filtering method to the ore image,so as to reduce the noises and keep the edge of the particles.(2)To solve adhesion particles cannot be effective segmentation.It is proposed that Distance Transform and Seed for Integral Image(DTSI)method to segment.Through the comparative analysis with several common segmentation methods,the results show that this algorithm can accurately divide the adhesive particles into individual particles.(3)According to the actual measurement and conversion relationship of pixels,calculating the actual parameters of each particle in the ore image characteristics,demarcating the particle segmented regions,on the basis of the statistics of the number of particles to draw the accumulated percentage particle number and the particle size distribution curve.