Research And Realization Identification And Positioning Method Of Coal And Gangue Based On Image

Sorting coal and gangue is a necessary process in coal production,which can improve coal mine safety,upgrade coal grade,reduce washing costs,and improve the economic benefit of enterprises.Currently,the sorting of coal and gangue mainly relies on manual sorting and mechanical sorting.These two methods are labor-intensive,consume a large amount of energy,and cause environmental pollution.Therefore,as a new coal preparation equipment,the coal and gangue sorting robot has attracted extensive attention in the industry.Identification and positioning of coal and gangue are the primary parts of the sorting robot,which directly affects th e sorting effect.With the development of image processing technology,the identification and positioning of coal and gangue from the perspective of image analysis or visual calculation has attracted the attention of researchers.This paper mainly carries out the following work:(1)According to relevant regulations of coal production and the actual production environment,a simulation experiment platform is built in the laboratory to obtain sample images and establish a single sample image database.The experiment is conducted to compare three kinds of filters for noise reduction of the images,which indicates that the nonlinear low pass filtering achieves the best performance.Considering that the surfaces of coal and gangue differentiate in grayscale and texture,the features of the grayscale and texture of the samples are analyzed,it is found that coal and gangue have higher discrimination degrees in gray variance,skewness,texture contrast,and entropy.(2)K-Nearest Neighbor and least squares support vector machine are selected as image recognition algorithms for coal and gangue.Three classifiers are trained by using the feature of grayscale,the feature of texture,and the joint feature combining skewness with contrast respectively.The experimental results show that the classifier using the joint feature has the best result,and the algorithm of coal and gangue recognition based on the least square support vector machine is more accurate.(3)Three automatic image threshold segmentation methods are used for image binarization processing.The comparative analysis shows that the clustering method had the best result.Particle analysis of coal and gangue samples is carried out by morphological corrosion and expansion methods to obtain a complete and clean target sample.The overlapping samples were separated and reconstructed by morphological corrosion and expansion methods,and the center of mass is extracted by using the center of mass method in the coal-gangue mixing image.The camera is calibrated and the mapping relationship between the real coordinate system and the pixel coordinate system is obtained.(4)Based on the LABVIEW platform developed coal gangue image recognition and positioning system,the system mainly includes image acquisition,image filtering,image feature extraction,sample training,sample classification,image binarization,background removal and filling and center of mass extraction.The system realizes the automatic acquisition and processing of coal and gangue images,the analysis of grayscale and texture features,and the identification and positioning of coal and gangue.(5)Set up an experimental platform,the positioning accuracy,identification accuracy,and system running speed of the image identification and positioning system of the coal and gangue sorting robot are tested with randomly selected coal and gangue under actual working conditions.The test results show that the average coordinate error of positioning in the X direction is 5.0mm and that in the Y direction is 4.7mm.Different types of light sources,light intensity,and surface humidity of coal and gangue all have certain influences on the identification results of coal and gangue,when the light source type is LED strip lamp and the light intensity is 3001ux,the identification accuracy of coal and gangue samples is 88.3%and 90.0%respectively.The sum of time of identification,positioning,and opening the camera for a single sample on average is 0.130s.Therefore,this paper studies the method and system of coal and gangue identification and positioning,which can realize the automatic identification and positioning of coal and gangue more accurately.