Research On Particle Image Recognition And Its Application In Blast Furnace Slag Particle Size Detection

In view of the low efficiency of energy recovery and resource utilization of blast furnace slag in the iron and steel industry,researchers at home and abroad have proposed a process for mechanical centrifugal granulation of high-temperature liquid slag,which requires that the proportion of blast furnace slag particles with a particle size of less than 2 mm after granulation is more than 95%.In order to achieve this technological requirement,the particle size of the blast furnace slag in the granulation process needs to be detected in real time,and then a detection method for image recognition is proposed.Based on this method,on the VS2010 and HALCON software platforms,the hardware and software system of the "blast furnace slag particle size detection system" was developed,and the system was tested in cold state and hot state.After further debugging,it was realized real-time detection of blast furnace slag particle size.The main contents of the study are as follows:(1)According to the characteristics of the obtained blast furnace slag cold particles,the theoretical algorithm of image processing is studied.In the process of image filtering,according to the type of noise that may occur in the granulation process,a denoising algorithm based on the adaptive median filter and wavelet threshold is proposed.Experimental results show that the algorithm has a good denoising effect.As for the binarization process,in the problem of the difference in the gray characteristics of the particles,on the basis of the integral image theory,the adaptive local threshold algorithm is used to binarize it.The experimental results show that compared with the traditional binarization algorithm,the image edge is better protected.As for the blast furnace slag particles sticking to each other in the image,the improved watershed algorithm is used to segment it,and the roundness threshold is set to filter the particles.Compared with the classic watershed algorithm,this algorithm suppresses the over-segmentation phenomenon to a certain extent,and improves the accuracy of the segmentation.(2)By analyzing the function and structure of the system,the system detection scheme was researched and designed,and a blast furnace slag particle size detection system was developed.According to the requirements of system detection accuracy and grain extraction range,the optical equipment is selected and its installation device is designed.In order to collect the high-temperature slag particles in the granulator,the slag collection mechanism and its control system are designed and developed;Based on the image processing theoretical algorithm,on the basis of the VS2010 software platform and HALCON machine vision algorithm library,research designed the "blast furnace slag particle size detection system" interface,and calibrated the system through the standard parts of the steel ball.(3)The experimental effect of the developed blast furnace slag particle size detection system was examined.Using common round-like particles,such as soybeans,red beans,mung beans,black beans and blast furnace slag cold particles,the test results show that the system can realize the particle size detection of various round-like particles within a relative ± 3.5 % error.In order to further verify the particle size detection effect of the system in the centrifugal granulation process of high-temperature liquid slag,the blast furnace slag particles are detected by sampling.The blast furnace slag particles are collected by the slag taking device,and the system performs online detection and detection is done.The results show that the relative error of the experiment is less than 4.1%,and the system can meet the design requirements and realize the real-time detection of the particle size during the high-temperature liquid slag centrifugal granulation process.In summary,the researched blast furnace slag particle size detection method based on image recognition is feasible,and the developed blast furnace slag particle recognition system can also meet the corresponding system requirements.