基于改进超像素分割算法的粘连岩石分割  

Segmentation of Cohesive Rock Based on Improved Superpixel Segmentation Algorithm

作　　者：李承阳 沈建琪[1] 

机构地区：[1]上海理工大学,上海

出　　处：《应用数学进展》2022年第6期3903-3908,共6页Advances in Applied Mathematics

摘　　要：碎矿石的粒度信息是对选矿最优控制的主要指标,为了能够快速准确地检测矿石粒度,提出了一种结合凹点检测和超像素分割的粘连岩石分割方法,通过凹点数量来确定超像素块的个数直接对粘连区域进行分割。研究结果表明,本文算法速度快,准确率也比传统的超像素分割高。Particle size information of crushed mineral is one of the main factors for optimal control in mineral processing. In order to measure the particle size of mineral quickly and accurately, an adhesive rock segmentation method combining concave detection and super-pixel segmentation is proposed. The number of concave points is used to determine the number of super-pixel blocks to segment the adhesive area directly. The results show that the proposed algorithm is faster and more accu-rate than traditional superpixel segmentation.

关 键 词：粘连分割 凹点检测 超像素分割 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]