Current Trends and Future Research Direction in Automatic Cracks Detection and Interpretation

Sara Ashraf ; Islam Hegazy ; Taha Elarif 


Abstract


The periodically checks and maintenance of structures is a critical step. Older than twenty years ago those checks was done manually, where specialist observe surface material and make analysis. However, the output of manual method always differs from one specialist to another. Thus, the manual method become unpractical, which motivates researchers to automate the check-up process. Digital image processing techniques for crack extraction are already widely implemented on large highway maintenance projects. In this paper, an overview is conducted to explain the automatic crack detection systems life cycle, concrete surfaces distress types, best-acting techniques is listed, and the performance measures. Followed by the challenges and open research points. Finally, present a proposed future work on this field. The challenges still opened on this field, beginning by selecting the right portable device for automatic inspection to give a high quality image for processing, can inspect on hard weather and can access danger places. Then, at pre-processing step, there exist many image noises have to be neglected to output accurate cracks; also, adapting lights is crucial, because on dark images, the cracks become unclear to be distinguished from background surface. Then, the detection process is depending directly on the previous pre-processing step, some algorithms still needs many trails for parameters adaptation. However, the Error-nous rates is decremented now a day with the improvement of crack detection algorithms. Later, the crack characterization with severity identification still needs more research attention to full automate systems and make decisions without any human intervention.


Other data

Keywords Crack Detection, Image processing, Crack Interpretation, Concrete Surface Distresses.
Issue Date Jul-2019
Publisher ITHEA
Journal International Journal "Information Models and Analyses" 
URI http://research.asu.edu.eg/handle/123456789/170305


Recommend this item

CORE Recommender

Items in Ain Shams Scholar are protected by copyright, with all rights reserved, unless otherwise indicated.