An Image Analysis Based Damage Classification Methodology

Measurement of the extent of damage in a real structure is extremely important in terms of any maintenance strategy. However, this measurement often turns out to be difficult, time consuming and error – prone. The necessity of a simple, fast and relatively inexpensive damage monitoring system with reliable measurements is growing for quite sometime. The paper proposes a camera based image analysis technique to quantify and classify damage in structures at various levels of scale. The general method has been applied to corroded plate specimens in the laboratory with the aim to identify the affected areas on a steel pile due to pitting corrosion. The method depends on the contrast of the corroded region with respect to its surroundings, performs intelligent edge detection through image processing techniques and computes each affected and closed region to predict the total area of the affected part along with its spatial distribution on a two dimensional plane. Moreover the performance of the camera allows defining a detection threshold and the so-called probability of detection (PoD) and probability of false alarms (PFA). PoD are suggested as functions of the area of the pitting for the construction of Receiver-Operating-Characteristic (ROC) curves. The methodology can be used as a tool for the owners/managers of the structure for objectively quantifying and localising the extent of pitting corrosion, rather than providing information through a subjective visual assessment. Moreover, it allows introducing the probability of detection and probability of false alarms in the decision chain and in risk analysis. The method is shown to be robust, reliable, simple and inexpensive