CAJ | 학술논문

在大型古建筑物测量的过程中，由于残缺区域边界纹理特征分布极为复杂，造成残缺区域特征在图像重构的过程精确度降低，导致难以进行准确测量。提出一种基于边界特征自动提取算法的大型古建筑残缺区域施工测量方法。根据残缺区域边界的曲率计算结果，结合种子格栅选取和自动区域生长等方法将大型古建筑残缺区域的关键数据点从背景中进行有效分离，根据多边形细分和离散光顺相关原理对大型古建筑残缺区域的边界进行重构，提高了大型古建筑测量的准确度。试验结果表明，利用边界特征自动提取算法能够有效将大型古建筑残缺区域进行重构，极大地提高了大型古建筑测量的准确度。
재대형고건축물측량적과정중，유우잔결구역변계문리특정분포겁위복잡，조성잔결구역특정재도상중구적과정정학도강저，도치난이진행준학측량。제출일충기우변계특정자동제취산법적대형고건축잔결구역시공측량방법。근거잔결구역변계적곡솔계산결과，결합충자격책선취화자동구역생장등방법장대형고건축잔결구역적관건수거점종배경중진행유효분리，근거다변형세분화리산광순상관원리대대형고건축잔결구역적변계진행중구，제고료대형고건축측량적준학도。시험결과표명，이용변계특정자동제취산법능구유효장대형고건축잔결구역진행중구，겁대지제고료대형고건축측량적준학도。
In the process of measuring large ancient buildings, the damaged area boundary texture feature distribution is very complex, and it leads to reducing of image reconstruction accuracy for the imperfect regional characteristics, making it difficult to accurately measure. For this reasons, this paper puts forward a kind of measurement method of large buildings damaged area based on boundary features automatic extraction algorithm. According to the curvature of the damaged area boundary estimation results, the key data points are effectively separated from the background combined with seed grating selection and region growing method to large ancient buildings damaged area. The damage area edge in large ancient buildings is reconstructed according to the principle of polygon subdivision and discrete fairing, so as to improve the measurement accuracy of large ancient buildings. The experimental results show that the boundary characteristics of automatic extraction algorithm can effectively reconstruct the large ancient buildings damaged area, and it greatly improves the measurement accuracy.