A study on human-machine collaborative software for making a defects map of concrete structures

2017 Fiscal Year Final Research Report

• Project/Area Number: 15K06180
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Research Field: Structural engineering/Earthquake engineering/Maintenance management engineering
• Research Institution: Yamaguchi University
• Principal Investigator: Kawamura Kei 山口大学, 大学院創成科学研究科, 准教授 (70397991)
• Project Period (FY): 2015-04-01 – 2018-03-31
• Keywords: 維持管理 / ひび割れ抽出 / 画像処理 / 社会基盤 / コンクリート施設 / 知的情報処理 / 対話型遺伝的アルゴリズム

Outline of Final Research Achievements

This study proposed a semiautomatic method for concrete surface cracks extraction from digital images. The characteristic feature of the method was the combination of interactive genetic algorithm (iGA) and touchscreen. The iGA was applied to optimize image processing parameters for each image without knowledge of image processing techniques. The touchscreen enabled the user to touch images directly and extract cracks smoothly. In order to show the capability of the software which implemented the proposed method, concrete surface cracks were extracted from digital images with reasonable accuracy and processing time.

Free Research Field

維持管理工学