Eureka! Institutional Repository

Cracks in concrete cement exist in all constructions. Their development is commonly desirable since they indicate an existing problem. Determining the quantity of cracks and surveillancing the severity of them is of outmost importance regarding the current condition of the concrete and for determining the lifespan of a construction. Cracks in the concrete may be an indication of corrosion in the metallic armoring, shard development, or alteration in the shuttering conditions. Thus, the surveillance of cracks throughout the lifespan of concrete-based constructions is an efficient technique to evaluate the safety level of the construction itself and for planning future restoration/repair. A newly introduced technique for surveillance is the use of unmanned aerial vehicles (UAV). UAVs are drones equipped with cameras, sensors, GPS, etc. RGB pictures (colored pictures in red, green, and blue tones) are captured by cameras loaded in the UAV, which flies around the construction, seeking to locate cracks and other defects. Each picture taken by the UAV should be evaluated to located cracks. For timesaving purposes, the evaluation can be done by using image processing techniques for auto-detecting cracks and defects, rather than using human optical inspection to locate them. The current work is organized as follows: An introduction on how computers understand and process images is presented in chapter 1. The proposed method for autodetecting cracks is introduced in chapter 2. Experimental results of the proposed method are shown in chapter 3. Future expansions and inferences regarding the current work are concluded in chapter 4. Finally, in the appendix, it is presented the Matlab code used