Abstract: Optimizing blasting schemes based on the fragmentation distribution characteristics of muck piles is crucial for
improving production efficiency in open-pit mining. However,theoretical models are susceptible to on-site factors,leading to deviations,
while 3D laser scanning has limitations such as complex procedures,high cost,and significant risks. Therefore,taking
the Mo′ergou open-pit limestone mine in Sichuan as an example,this study used UAV oblique photography to obtain multi-angle
muck pile image data and proposed a fragmentation prediction method based on a 3D muck pile model. To address the issue
that particle size analysis of muck pile surface images cannot reflect the internal distribution,a distribution coefficient was introduced
to correct the difference in fragmentation distribution between the surface and the interior,thereby estimating the overall
fragmentation distribution of the muck pile,and the corrected results were validated. Finally,using the boulder yield as an evaluation
index for blasting effectiveness,the size threshold for boulders was defined based on actual production requirements,
and the boulder yield of blasted rock was predicted. The results show that:① The constructed 3D model meets the mapping accuracy
requirement of 1∶500,and the average error between the corrected overall fragmentation distribution and the measured
results is only 4. 26%;② Rocks and ores with a side length greater than 80 cm are defined as boulders,and the measured
boulder yield is 3. 85%,meeting the mine′s production requirements. The proposed method can provide a useful reference for
evaluating blasting effectiveness in open-pit mines.

Key words: open-pit mine,UAV oblique photography,three-dimensional model,particle size analysis,bulk rate