Research And Analysis Of Ore Granularity Image

With the continuous development of computer vision technology in recent years,image processing technology has become more and more widely used in various fields.Image processing technology can improve the traditional industrial production methods and realize the development of industrial intelligence.It is "Made in China 2025" An important breakthrough point.Ore particle size is an important basis for measuring the degree of ore fragmentation,and it is also an important parameter affecting the selection of ore dressing methods and process flow.Traditional ore particle size detection is done manually.This method is not only time-consuming and labor-intensive,but also has low efficiency.Using image processing-related technologies,real-time detection of ore particle size can be achieved,which greatly reduces labor costs and improves detection efficiency and accuracy.Therefore,the ore particle size detection based on image processing technology has great development prospects and research significance.In order to realize the online automatic detection of ore granularity,this paper uses an image processing-based method.In the image processing process,two difficulties need to be solved: denoising and segmentation of the ore image.In this paper,a new lifting wavelet construction method is proposed: lifting the wavelet based on the cubic B-spline function to achieve denoising of the ore image without destroying the grain information of the ore image.For the ore image segmentation,this paper proposes an improved watershed algorithm.The specific steps of ore particle size detection are: first perform a grayscale transformation on the ore image to reduce the amount of computation,and use a lifting wavelet transform based on the cubic B-spline function on the grayscale image to decompose the ore image into four sub-pictures,and then Perform threshold denoising processing on the four sub-pictures,and finally perform lifting wavelet reconstruction to obtain the denoised ore image;then use the improved watershed algorithm to segment the ore image.The step of the improved watershed algorithm is to use morphological reconstruction,Mark the foreground and background of the ore image,remove the irrelevant extreme values in the gradient image,and then use the watershed algorithm to get the ridgeline map of the ore image;then use the connected domain properties of the image to calculate the corresponding ore particle The number of connected domains and the number of pixels in each connected domain,and then the conversion relationship between the number of pixels and the actual ore size is used to obtain the ore particle size.In order to verify the accuracy of image processing to detect the ore particle size,the ore particles need to be manually screened to calculate the actual ore particle size and compared with the image processing results.The comparison result is that the maximum error between the two is 3%.The feasibility of image processing to detect the particle size of ore is presented.