【摘要】 为了实现矿石粒度的在线自动化检测，需要解决两个难点：矿石图像的去噪和分割。通过提出了一种新的提升小波构造方法：基于三次B样条函数的提升小波，实现了对矿石图像的去噪。对于矿石图像的分割，提出了改进的分水岭算法。矿石粒度检测的具体步骤是：首先利用基于三次B样条函数的提升小波对图像进行去噪，再利用改进的分水岭算法对矿石图像进行分割。最后利用图像的连通域性质，计算各个连通域的像素面积，再转换到实际的矿石粒度大小，从而实现对矿石粒度的检测。对比这里算法与人工筛选的结果，累积误差在3%以内，可见这里算法具有可行性和准确性。更多还原

【Abstract】 In order to achieve online automated detection of ore particle size,There are two difficulties need to be solved:denoising and segmentation of ore images. A new lifting wavelet construction method is proposed:the lifting wavelet based on the cubic B-spline function is used to denoise the ore image. For the segmentation of ore images,it proposes an improved watershed algorithm. The specific steps of ore particle size detection are:firstly denoising the image by lifting wavelet based on cubic B-spline function,and then using the improved watershed algorithm to segment the ore image.Finally,using the connected domain properties of the image,the pixel area of each connected domain is calculated,and then converted to the actual ore size,thereby realizing the detection of ore particle size.Comparing the results of the algorithm and manual screening,the cumulative error is less than 3%. It can be seen that the proposed algorithm is feasible and accurate.更多还原