Abstract Angle of repose (AOR) is one of the essential parameters to study the flowability of granular materials. It is also commonly used to calibrate the parameters of discrete element method (DEM). In practice, AOR is measured manually or using digital image processing (DIP). However, reliable and consistent measurement of AOR is still a challenging task due to the irregular or nonlinear shape of the material pile, particularly for the AOR of cohesive materials like soils. This study suggests an accurate and consistent way to measure AOR using DIP. The developed algorithm automatically finds the AOR of the material pile by a search algorithm for the unique maximal least square using linear regression analysis. The proposed algorithm is validated using customised templates with known angles, followed by an evaluation study against (a) an existing DIP method using a linear regression within “manually” selected region of interest and (b) the results of manual measurements using a protractor. The results showed that the developed algorithm can accurately measure and quantify AOR of soil piles. It was concluded from the results that the developed algorithm can be used to accurately measure AOR without error-prone subjective judgement.

中文翻译：

---

一种基于数字图像处理的最大最小二乘法自动测量颗粒材料休止角的新方法

摘要 休止角(AOR)是研究颗粒材料流动性的重要参数之一。它也常用于校准离散元法（DEM）的参数。在实践中，AOR 是手动或使用数字图像处理 (DIP) 测量的。然而，由于材料堆的不规则或非线性形状，可靠和一致的 AOR 测量仍然是一项具有挑战性的任务，特别是对于土壤等粘性材料的 AOR。这项研究提出了一种使用 DIP 测量 AOR 的准确且一致的方法。开发的算法通过使用线性回归分析的唯一最大最小二乘法的搜索算法自动找到材料堆的 AOR。使用具有已知角度的定制模板来验证所提出的算法，随后是针对 (a) 在“手动”选择的感兴趣区域内使用线性回归的现有 DIP 方法和 (b) 使用量角器手动测量的结果的评估研究。结果表明，所开发的算法可以准确地测量和量化土桩的 AOR。结果表明，所开发的算法可用于准确测量 AOR，而不会产生容易出错的主观判断。