Localization of an unmanned aerial vehicle for crack detection in railway tracks

An Unmanned Aerial Vehicle (UAV) has several applications in the modern scenario. It can be used for capturing live video footages for sports like cricket or football or even commercial applications. It can even be used for procuring images at places where human intervention is difficult and so, can find wide application in disaster management. This paper mainly deals with the localization of a UAV and how it can be applied for detecting cracks in a railway track using the concepts of image processing. The algorithm used for localizing the UAV is called Monte Carlo or Particle filter localization algorithm. The video taken from the UAV is processed by further extracting image frames and analyzing each of these image frames for detection of cracks along railway line. A real-time implementation of our proposed method can significantly reduce physical labor involved in crack detection and also reduces the risk of accidents.