Research And Design Of Online Particle Size Detection System Based On Image Recognition

The particle size information of crushed ore reflects the working status of crushers at all levels in the crushing process.The particle size distribution of ore can be mastered in real time by online particle size detection.The test results are fed back to the crusher control system,which can effectively control the working parameters of the crusher.So as to improve the efficiency of mineral processing,reduce energy consumption,improve the utilization rate of mineral resources.Online particle size detection technology based on computer vision image processing,which has many advantages such as safe and fast non-contact measurement,is a hot spot of current particle detection technology research.Aiming at the spot production environment,the image processing technology is used and the on-line granularity measurement system is designed in this research.The main results of this research are as follows:(1)The resolution algorithm(NRSS)based on gradient structural similarity is used to compare the ore images with different resolution and content.A preliminary NRSS threshold that can be used to determine whether an image is clear.Finally,the validity of threshold selection is verified by randomly collected images.The results show that the threshold can accurately evaluate the sharpness of collected images.(2)Aiming at the problem that the adhesive particles in the ore image cannot be effectively segmented,an improved watershed transform segmentation method based on multi-scale morphological gradient is proposed.Compared with the common segmentation algorithms,the results show that the particle target can accurately be segmented in this segmentation method.(3)According to the relationship between the actual measurement unit and pixel calibration,the actual parameter characteristics of each particle in the ore image are calculated.The ore segmentation region is demarcated.The ore particle size distribution is statistically analyzed.And the parameter distribution map and particle size distribution map are drawn.The results show that the difference between the detected particle size distribution results and the actual screening results is small and has high accuracy.(4)By combining MATLAB image processing toolbox and GUI development environment(GUIDE),the software part of granularity online detection system is established.And the system operation interface is designed.The communication between MATLAB and Excel is used to realize the function of image loading,image denoising,image binarization,image segmentation,particle size detection and particle size data preservation.The ore image can precisely be segmented and the on-line particle size detection is realized by the on-line particle size detection system that is designed in this research.