JpGU-AGU Joint Meeting 2020

講演情報

[J] 口頭発表

セッション記号 M (領域外・複数領域) » M-IS ジョイント

[M-IS28] 歴史学×地球惑星科学

コンビーナ:加納 靖之(東京大学地震研究所)、磯部 洋明(京都市立芸術大学美術学部)、芳村 圭(東京大学生産技術研究所)、岩橋 清美(国文学研究資料館)

[MIS28-06] 碑文解読のための画像処理アプリケーション開発

*上椙 英之1上椙 真之2多仁 照廣3 (1.独立行政法人 国立文化財機構 奈良文化財研究所、2.公益財団法人 高輝度光科学研究センター、3.若狭路文化研究所)

キーワード:自然災害伝承碑、画像処理、碑文

There are many stone monuments in the areas affected by the Tokai and Nankai earthquakes in the past. However, most stone monuments have deteriorated because of weathering. Furthermore, accurate information is no longer available if the stone monuments themselves were lost. Therefore, it is urgently necessary to archive as many stone monuments as possible.

It is difficult to decipher weathered inscription from ordinary photographs. Due to the ambiguity of the photo images of inscription, digitizing and archiving the text on the surface of stone monuments is always problematic. The quality of the images taken under different conditions changes largely, making it also decrease the quality of the archive.

Recently, methods for restoring the shape of stone monuments based on the images taken from different direction have dramatically advanced, and it became possible to automatically restore the three-dimensional shape from a large number of digital photographs [1-2]. However, it was difficult to apply those methods to the fieldwork of the huge number of stone monuments because of complex and time-consuming procedure of the methods.

We are developing a new method of digital archiving of inscriptions [3-4]. Our method emphasizes three features, “Speedy”, “Simple”, and “Lightweight” for the fieldwork, because of the variable condition of stone monuments. Photographic instruments should be made lightweight to make it possible to visit several places with variable conditions. For this reason, we constructed the new photography scheme as simple as possible in the beginning of the series of our studies [reference].

We also developed image processing software for digital archiving of inscription on stone monuments. There are two features in the software. First, shadow of the engraved characters on the stone monument was enhanced for easy recognition of text by removing background noises. Second, the area only illuminated by oblique hand-light was extracted by masking with the illumination image.

The software succeeded in speeding up for the procedures of image processing step, and also improved the readability of the inscription text. Our new scheme and software can applicable not only inscriptions, but also divers samples of the historical and Archaeological materials, to digitize their surface shape in speedy.



REFERENCES:

[1] Koutsoudis, Anestis, et al. (2014) Multi-image 3D reconstruction data evaluation, Journal of Cultural Heritage 15.1: 73-79.

[2] Kennett, Douglas J., et al. (2012) Development and disintegration of Maya political systems in response to climate change, Science 338.6108: 788-791.

[3] Uesugi, Hideyuki, and Masayuki Uesugi. "Development of the Database for Images of the Text on the Stone Monuments." 2013 International Conference on Culture and Computing. IEEE, 2013.

[4] Uesugi, Hideyuki, Masayuki Uesugi, and Teruhiro Tani. "Image Processing Scheme for Archiving Epigraphs." 2018 3rd Digital Heritage International Congress (DigitalHERITAGE) held jointly with 2018 24th International Conference on Virtual Systems & Multimedia (VSMM 2018). IEEE, 2018.