Environmental factors affecting concrete structures like bridges, beams, columns and highways in onshore and offshore environments lead to development of micro-cracks. Early detection of surface micro-cracks in concrete structures helps to put preventive measures in place to avoid failure potentially saving loss of assets and in some cases lives. In the last decade alone image analysis-based methods have gained a lot of momentum as early non-invasive crack detection methods. More recently artificial intelligence-based methods, using ANN and CNN type neural networks, have also been applied to automatically process the images for crack identification. Although all these methods claim very high accuracy, they often ignore the complexity of the image collection process itself. Most of the published literature deals with images in ideal laboratory based conditions. Real time images are often impacted by illumination conditions, randomness of crack shapes, and irregular size of cracks, and various noises such as shadows, shading, blemishes and concrete spall in the acquired images. Testing of published image analysis methodologies on real world images of concrete structures often results in misleading results. In this project an automated AI based system will be created, which will allow training and testing of real time images of concrete bridges and offshore structures, which are augmented by the presence of shadows and other noises. In this project significant effort will be spent on the development of an image database of images of concrete structures with cracks and shadows, which will then be used for training and testing of the AI network created specifically for this project.
Project funding:
KTU Research and Innovation Fund
Project results:
As a result of the implementation of the project, two articles were published in international scientific journals, two were presented at international conferences, applications for continuing projects with foreign partners were prepared, real experimental data sets were exchanged, and detailed scientific research was carried out with them.
Period of project implementation: 2021-05-18 - 2021-12-31
Project coordinator: Kaunas University of Technology
