Adham Mohamed, Ahmad Mostafa, Mohamed Mahmoud, Mohamed Saed

Supervised By: Professor Alaa Hamdy, Eng. Hager Sobeah, Eng. Mahmoud ElSahhar

Publishing Date

27th of October 2020


Concrete pillars are designed to withstand different environmental conditions like earthquakes and to carry tons of loads. Buildings have multiple reinforced concrete columns to carry them and protect them. but, if one of them got damaged, the whole building would lose its integrity causing the building to fall apart costing lives and wasting huge amounts of money. The main idea of the project is identifying defects in concrete by scanning the pillars of the building where the defects like (voids or fractures) might exist or not, to do so we would need to scan the inside of the concrete to acquire these defects that are why engineers and users use the GPR (Ground penetration radar) in the collecting of the images because these defects are impeded within the concrete, after the acquiring phase then the user or the engineer will use his prediction to determine the defect which has a high human error factor because it`s only based over the experience of the user which higher the risk factor even more. Image processing and deep learning techniques will be used to detect and classify those defects according to their types using the VGG-16 algorithm.

1.1 Background

A large Number of constructed buildings are in increase daily worldwide especially in China, India & the United state as they are considered as top three countries in constructions, as earlier before 20th-century people suffered from the building’s defects as it ends horribly as the owners consider that their buildings are durable as it lasts for a long time as the structure segments break down at different rates and degrees depending on the plan, materials, and techniques for development, nature of workmanship, ecological conditions and the employments of the structure, Imperfections result from the reformist crumbling of the different segments that make up a structure. Deformities happen through the activity of one or a blend of neutral factors, and those problems ended with horrible disasters till the mid of 20th century as a lot of scientists discovered that most of the common mistakes that happen in the building structures are human mistake and they also said that the neutral factors doesn’t effect that much on the structure of the buildings like the human mistake for example , Argyris and Schön in 1978 they analysed more than 2000 document error from building and civil engineering projects , The purpose was to understand how defects occurred and most of the defects were human mistake , At the end of 20Th century Australia published the first building standard inspector in 1997 and it was one of the first solutions but it couldn’t be the perfect solution as it was only detecting the visible defects like cracks and stains but no one had any other solution until 2007 the inspectors started using GPR(Ground Penetration Radar) images to identify the defects that cant be seen it worked for a long term but wasn’t accurate enough as it depended on the experience of the inspector to identify if its considered as defect or not and it costed a lot of money and time to identify and study the defects that cant be seen and in this period the disaster might happen , and by the time the detecting the defects became easier as algorithms became provided that helped them in detecting the errors but till now there isn’t any accurate method made that can detect the percentage of the defects correctly , now we combined those methods that had been used till now to provide an easy and accurate way by using both of machine learning and image processing.

1.2 Motivation

Over the past years, different approaches have been proposed on the way designing concrete structures like building, Bridges and dams. With complex designs of buildings and the use of new materials like steel, concrete constructions’ durability has been affected due to lack of correct monitoring procedures and care.

Mix design, mechanical overload, chemical attack, poor construction, mechanical overloading and environmental exposure are some factors that affects concrete strength and durability. Those factors affect concrete and cause some defects; voids, corrosion, cracks, water leakage, deterioration. Those defects

cause sever damage to the integrity of concrete causing building and bridges to fall leading to a serious amounts of injuries and an increase in death rates, According to statistics ,And the most resent solution was to use the GPR B-Scan images and by an engineer trained eyes for more over 4 months he would be able to detect it and not with a high accuracy , Our solution would be to train our own module so that he can have a system witch would help him archiving higher accuracy and taking actions .

1.3 Problem Statement

1-distinguish between negative media ( wall) and positive media (Concrete) by training GPR B SCAN images of pillars concrete using machine learning.

2-Detecting Concrete’s defects in early stages by training the GPR images using machine learning and image processing (Histogram).