Algorithm for Automatic Crack Analysis and Severity Identification

Concrete structures are increasing every day, to facilitate people's lives. With this expansion, the traditional manual maintenance method becomes unpractical, costly and time-wasting. The fast detection and maintenance of concrete surfaces defects is necessary to save people's lives, reducing maintenance cost, and increase the lifetime of concrete structures. Thus, the researches came up over the last twenty years to find an automatic way in order to maintain, apply regularly check-ups over concrete structures and assist engineers to take fast decisions. The most researches came up with high precision algorithms to allocate cracks and defects over the concrete surfaces with no human intervention. Nowadays, the computer programs can be dependable to capture large data sets of concrete structure, and then give precise locations of cracks. However, there exists a lack of researches that work on crack interpretation and automatic decision-making, which is considered as a critical part of those systems. Therefore, there exists a need for methods that describe the crack characteristics in terms of width, length, and other morphological attributes. In this paper, a crack interpretation algorithm is proposed to extract crack geometrical attributes and support the decision maker.