品質管理 (1)

[VI-86] Basic examination on the quality judgment technique of concrete joint by machine learning

*中村 真人¹、松本 修治¹、今井 道男¹、柳井 修司¹、有坂 壮平¹ (1. 鹿島建設株式会社)

Keywords:Concreate, Construction joint, Histgram, Machine learning

The quality of the treatment on the joint surface of the concrete structure is often judged by visual inspection of the workers and supervisors, and it is rarely evaluated quantitatively. The authors developed a technique for judging the quality of joining surface focusing on the luminance distribution of images that can be used with tablet PCs did. However, in the technology that uses the luminance distribution, stable evaluation is difficult because the shooting conditions of the picture affect the judgment. In this paper, we show the result of the basic examination of the pass / fail surface judgment technology using machine learning.