ABSTRACT
This paper proposed the UAV-based crack identification using the image processing approach. The system hardware consisted of UAV, camera, ultrasonic displacement sensor, WiFi module, and Raspberry Pi for sensing and communication controller, being capable of providing crack images with associated distances on demand. The image processing strategy tailored to the crack images from the UAV-based system was proposed for the crack identification. The obtained crack images with the distance measurements were deblurred if necessary, subtracted using the median filter to enhance the crack areas, and binarized to separate them, which led to the crack width calculation. The combination of the hardware system and the image processing strategy allows the crack identification for full-scale structures.
