非破壊検査・診断Ⅱ

[1339] A STUDY ON SEMI-AUTOMATIC CONCRETE CRACKS DETECTION USING INTERACTIVE GENETIC ALGORITHM

Cuong NGUYEN¹, 河村圭², Amir TARIGHAT³ (1.Yamaguchi University 2.山口大学 3.Shahid Rajaee Teacher Training University)

キーワード：crack detection, genetic algorithm, image processing, 遺伝的アルゴリズム, 画像処理, ひび割れ抽出

Concrete surface crack detection is an important task of the inspection and diagnosis of concrete structures. This paper proposes a prototype software for semi-automatic crack detection from concrete surface images using an interactive genetic algorithm (iGA) and touch screen. The iGA is applied to optimize image processing parameters for each image with no knowledge of image processing techniques. Furthermore, the experimental results indicate that the proposed method has possibility of extracting crack with reasonable accuracy and working time.