基于智能视觉的混凝土微结构质量检测方法研究

Research on Concrete Micro Structure Analysis Based on Intelligent Vision

摘要: 混凝土质量对建筑物，桥梁，公路等设施的安全性，可靠性影响很大。目前的质量检测主要是人工的方法，通过显微镜观察混凝土切面的成分组成和分布，费时费力，检测质量很难保证。本文提出一种新颖的基于模糊规则混凝土质量智能视觉识别技术，通过支持向量机自动学习模糊规则，模仿人类的智能视觉识别过程，提高质量检测的自动化，可靠性和效率。实验证明，该方法相对于传统的方法来说，是有效的。

Abstract: The quality of concrete plays an important role in assessing the safety and reliability issues of buildings, bridges and roads. At present, quality tests mainly depend on manual labor, through a micro telescope vision to check crossover of concrete sample. It is time-consuming and the test results have low reliability. This paper proposes a new intelligent vision based on analyzing method by implementing fuzzy logic. Base on support vector learning, the fuzzy rules are constructed automatically simulating a human learning and classifying process. This approach improves the productivity, reliability, and degree of automation. Compared with traditional method, this way proves its effectiveness through experimental verification.