<p>In the shipbuilding and offshore industries, dimensional errors are managed based on quality management points to ensure seamless block assembly. Terrestrial laser scanners are widely used to measure these points in the field. However, owing to the massive size of ship blocks, operators often rely on empirical judgment to determine scanning positions, resulting in an excessive number of viewpoints. This inefficiency prolongs both data acquisition and computational processing times. In this study, we propose a two-stage viewpoint optimization method that combines a genetic algorithm with a greedy algorithm. In the proposed method, initial viewpoints are randomly determined according to the number of quality management points, and a genetic algorithm is then used to determine the viewpoints that maximize coverage. Subsequently, a greedy algorithm is applied to select the minimum number of viewpoints while maintaining coverage. We conducted experiments on actual ship blocks by implementing the proposed method. The results show that selecting initial viewpoints using the genetic algorithm provided only a slight improvement in coverage compared to random selection. In contrast, the greedy algorithm reduced the number of viewpoints by approximately 48.65 % on average without any loss of coverage.</p>