Estimating Damaged Volumes on Three-Dimensional Models of Historic Pagodas in Bagan After Earthquake

Abstract

Natural disaster is one of the major threats to cultural heritage in the developing countries. Floods, earthquakes, fires, environmental fatigue or similar long-term climatic effects can cause irreversible damage to cultural heritage. It can completely destroy the entire cultural heritage. In Myanmar, a magnitude of 6.8 earthquake struck west of Chauk nearly Bagan from the depth of 84km on 24th August, 2016. This earthquake destroyed mainly the ancient pagodas in Bagan, Myanmar. Bagan is recognized one of the world heritage status in Asia. Nowadays there are many pagodas approximately 1000 pagodas in Bagan. Therefore the affected areas in historic pagodas were needed to perform the damage assessment on cultural heritage. This research presents an automatic detection of the damaged volume of the ancient pagodas after earthquake had struck in Bagan. This research emphasizes how many percentages of volume is damaged on the features of the 3D point clouds. 3D point clouds is a collection of data point with (x, y, z) position on space that are produced from different types of units such as 3D software, 3D scanner and so on. The 3D point clouds used in this research are got from the pix4D software. Pix4D software produces the 3D point clouds by operating the UAV images. The UAV images are acquired by taking the aerial photos using the hexacopter. The output point clouds of specific pagoda are mixed with other objects such as trees, other building and so on. The unrequired region needs to be eliminated and the region of interest of pagoda is extracted. The extracting of 3D damaged pagodas is mainly contributed in the proposed system. The proposed system gives the accurate noise removing for damaged pagoda. The information of pre-earthquake volume can be gotten from standardize 2D image to compare the post-earthquake volume. The damaged volume of the historic pagoda can easily be detected. The result for estimating top point of the damaged pagoda is compared with 3D Hough Transform of 3D point clouds. Thought the experimental results of the proposed system, it can be confirmed the proposed approach work adequately. It is hoped that this research will help the valuable cultural heritage of Myanmar to preserve. The effectiveness of the proposed system can be confirmed through the reasonable experimental outcomes of automatic damaged volume of theiv 3D historic pagoda. The system can be achieved with UAV images with minimum time consuming and cost.