Image Segmentation Technology And Image Recognition System Development Of Rockburst Based On Double Threshold

The blasting fragmentation is an important index to evaluate the blasting effect.It is of great practical significance to establish a fast detection and evaluation method of blasting fragmentation for mine production.In this paper,the computer image recognition technology is introduced into the information detection of the blasting block degree.According to the distribution characteristics of the large blasting block degree in the open pit mine,the image segmentation method of the blasting block degree is established,and the image recognition system of the blasting block degree is developed.The high-efficiency detection and quantitative feedback of large blasting block are realized.The specific research work and results are as follows.(1)Research on image segmentation technology of blasting fragmentation.In view of the characteristics of large-scale,conglutination,large difference and irregular shape of the open-pit mine blasting block,based on the comprehensive analysis of the adaptability of various image segmentation technologies,a double threshold image segmentation technology is developed.Through the comparison of different image segmentation techniques for blasting rock image segmentation,the developed dual threshold image segmentation method has advantages in the binary segmentation of rock block and background,and the filtering effect of rock surface noise is the best.The two threshold image segmentation method developed in this paper is used to segment the rock block images of different lithology,which shows that the established segmentationmethod has strong adaptability.(2)Development of image recognition system(bfas3.0).The process of image processing is designed.According to the characteristics of the collected image,the methods of image denoising and preprocessing of rock scene are proposed.For the characteristics that the uneven natural light causes the shadow in the target area and the gray histogram loses the double peak distribution,the preprocessing method of histogram equalization is adopted,and the methods of USM enhancement algorithm,morphological corrosion and expansion are applied to the gray scale of rock block The image is reprocessed.At the same time,a set of bfas3.0,which can quickly realize the block information display from the resolution derivation,image preprocessing,rock block location and segmentation,rock block geometry size calculation,is formed by testing the parameters according to the system framework.(3)Applied research.The application of bfas3.0 in a bench blasting in Huizhou,Guangdong Province was studied.First of all,an image acquisition system is established to collect images of 5 blasting piles with different blasting parameters.Then,bfas3.0 is applied to image import,resolution derivation,image preprocessing,image segmentation and finally to statistical results output,and the quantitative description index of each blasting block is obtained.According to the feedback results,the blasting parameters are optimized.