碎片(计算)
缩放比例
尺寸
片段(逻辑)
计算机科学
算法
样本量测定
统计
数学
几何学
操作系统
艺术
视觉艺术
作者
T.H. Kleine,A. R. Cameron
标识
DOI:10.1201/9780203747919-14
摘要
GoldSize is a Windows™ based computer program to estimate the sizes of objects in photographs. Principally its use to date has been in the estimation of blast fragmentation size distributions. Some features, such as fines correction to estimate the amount of material too small to be seen, are specifically tailored to fragmentation analysis. Also the sizes, by default, are empirically adjusted to match the results obtained when rock fragments are sieved.
In addition, the Kuz-Ram model for fragmentation prediction has been included in the program so that the measured fragment size distribution can be compared with the predicted size distribution and the fragmentation model parameters adjusted in the form of a calibration of the model. In this way a calibrated fragmentation prediction tool is combined with a fragmentation measurement tool which greatly improves the blast optimisation cycle.
The process of sizing is divided into four steps; obtaining photographs of the material to be sized such that a wide range of fragment sizes can be identified and a statistically acceptable sample of the material is obtained, object digitizing which is performed by tracing around the objects using the computer’s mouse, scaling the image which links the top and bottom of the image to the back and front of the collection of objects in the picture and the calculation of sizes based on object width.
Because it is not possible to size every fragment, fragmentation analysis is a statistical problem. It is thus necessary to include a statistical assessment of the results. The formulae are well established and may be used to calculate the reliability (or potential error) of the information about each size bin in the distribution.
科研通智能强力驱动
Strongly Powered by AbleSci AI