Casting surface specifications are set based on aesthetics, functionality, or a combination of both. To classify casting surfaces, visual inspections are performed by an operator who compares the casting surface to pictures or comparator plates (e.g., metal, plastic) that represent a certain roughness level. This inspection process is highly subjective, and disagreements arise on the acceptance of a casting between the casting producer and buyer. To minimize these disagreements and use developments in 3D scanning, this study aims to develop a digital surface characterization method. The method developed and implemented in this study utilizes underlying geometry estimation, abnormality detection, and a new roughness characterization formula based on a variogram to determine a surface roughness value. Tests were done to compare the new roughness characterization formula with existing quantification methods (i.e., Sa, Sq) and to compare the results of the method with human operators. The tests indicated that the variogram roughness was able to differentiate between the roughness levels of the current surface roughness standards GAR-C9 and SCRATA. In addition, the results are repeatable as well as reproducible and agree with operator judgment based on a ranking comparison between the operator and the digital method. Overall, the digital surface roughness method has the potential to improve the communication between casting suppliers and designers and make the surface roughness classification more reliable and repeatable.

铸件表面规格是根据美观性，功能性或两者的组合来设置的。为了对铸件表面进行分类，由操作员进行视觉检查，将铸件表面与代表一定粗糙度水平的图片或比较板（例如，金属，塑料）进行比较。该检查过程是高度主观的，并且在铸件生产者和购买者之间接受铸件时会产生分歧。为了最大程度地减少这些分歧并在3D扫描中使用开发，本研究旨在开发一种数字表面表征方法。在这项研究中开发和实施的方法利用了基础几何估计，异常检测以及基于变异函数的新粗糙度表征公式来确定表面粗糙度值。进行了测试，以将新的粗糙度表征公式与现有的量化方法（即Sa，Sq）进行比较，并将该方法的结果与人工操作人员进行比较。测试表明，变异函数粗糙度能够区分当前表面粗糙度标准GAR-C9和SCRATA的粗糙度水平。另外，结果是可重复的和可重复的，并且与基于操作员和数字方法之间的排名比较的操作员判断相符。总的来说，数字表面粗糙度法有可能改善铸件供应商和设计师之间的沟通，并使表面粗糙度分类更加可靠和可重复。平方），并将该方法的结果与人工运算符进行比较。测试表明，变异函数粗糙度能够区分当前表面粗糙度标准GAR-C9和SCRATA的粗糙度水平。此外，结果是可重复的，可再现的，并且与基于操作员和数字方法之间的排名比较的操作员判断相符。总的来说，数字表面粗糙度法有可能改善铸件供应商和设计师之间的沟通，并使表面粗糙度分类更可靠和可重复。平方）并将该方法的结果与人工算符进行比较。测试表明，变异函数粗糙度能够区分当前表面粗糙度标准GAR-C9和SCRATA的粗糙度水平。另外，结果是可重复的和可重复的，并且与基于操作员和数字方法之间的排名比较的操作员判断相符。总的来说，数字表面粗糙度法有可能改善铸件供应商和设计者之间的沟通，并使表面粗糙度分类更可靠和可重复。结果是可重复的，可重复的，并且与操作员基于操作员和数字方法之间的排名比较的判断相一致。总的来说，数字表面粗糙度法有可能改善铸件供应商和设计师之间的沟通，并使表面粗糙度分类更可靠和可重复。结果是可重复的，可重复的，并且与操作员基于操作员和数字方法之间的排名比较的判断相一致。总的来说，数字表面粗糙度法有可能改善铸件供应商和设计者之间的沟通，并使表面粗糙度分类更可靠和可重复。