Inductive thermography is a method of non-destructive testing, which performs well for detecting surface cracks on magnetic specimen. It is a non- destructive, fast and contact-free method; making it suitable for automated inspection and testing purposes. In this thesis inductive thermography is used for the detection and characterization of Head Checks. Head Checks are the result of a form of rolling contact fatigue commonly found in railway track. Static and scanning measurements were performed on reference pieces of rail, as well as rail pieces with real Head Checks. The data is acquired in a laboratory test stand and the measurements are evaluated by means of phase imaging. To this end a sequence of images was acquired, in a stopand-go motion. The phase imaging is achieved by calculating a pixel-wise Fourier transform of the recorded image sequence. Further routines were implemented in m-code to enable the rectification and registration of both the original and the phase images. The rectification is based on a metric obtained from the evaluation of a calibration image and now permits the measurement of distances in the infrared images as well as the proper registration of an image sequence. An edge-detection algorithm was used on the phase images, to automatically detect and characterize head checks from the phase image data. The sequence of phase images was finally merged, using registration, to obtain a panoramic view of the whole length of the piece of railway track.