Intensity correction and dynamic range expansion for digital x-ray radiograph images

Quantitative and qualitative analysis of NDE radiograph film is often accomplished using image processing tools. A prerequisite for such analysis is transformation of the data contained in the physical x-ray film into discrete numerical data suitable for image processing. Such transformation, called image acquisition or digitization, will introduce noise and spurious information (artifacts) into the original data. While certain qualitative data enhancement processes are tolerant of these data anomalies, significant errors may preclude quantitative analysis such as flaw sizing or flaw detection. In this paper, a method is presented to model some of the imperfections in the image acquisition step in an attempt to compensate their effects. In particular, techniques for compensation of intensity nonlinearities and dynamic range limitations are described