Improvement of crack detection probability by using magnetic camera and image processing

Magnetic camera consists of magnetic source, arrayed small magnetic sensors, magnetic lens, analog-to-digital converter and interface, computer and monitor. The quantitative magnetic field around crack and its processed results could be obtained by using magnetic camera, and the crack could be inspected and evaluated quantitatively. In addition, the magnetic camera has to uphold with large lift-off to protect sensors from weak environment such as high temperature and mechanical vibration. However, the sensor sensitivity would be decreased when the lift-off was increased. Correspondingly, the improved techniques are necessary for increasing sensitivity of magnetic camera and probability of crack detection at the large lift-off. This paper proposes an image processing method, which separates a global full scale to the several regions and repeats shadings in each region, to increase a crack detection probability in the magnetic camera images such as partial derivative B/partial derivative x and partial derivative B-2/partial derivative x partial derivative y.