Application of Tatian"s method to slanted-edge MTF measurement

The ISO 12233 method for the measurement of the spatial frequency response (SFR) of digital still cameras and scanners is based on the analysis of slanted-edge image features. The procedure applies a form of edge-gradient analysis to an estimated edge-spread function. As with all measurement, image noise can introduce bias error and variation into the resulting camera SFR and modulation transfer function, (MTF). It is often pointed out that applying a derivative filter to the estimated edge-spread function, as is done in the ISO method, amplifies this image noise. To reduce the influence of noise on the measurement, data averaging and fitting have been proposed. One method for edge-gradient analysis, reported by Tatian, avoids the above discrete derivative step. The MTF is expressed as a trigonometric series, whose elements are estimated from the measured edge-spread function. We describe the application of this method as an intermediate step in the ISO procedure. The method was benchmarked for both synthetic edges and captured test images. Results indicate good agreement between the two estimates for low-noise image data. For higher noise levels, however, Tatian's method is found to be susceptible to the selection (cropping) of the input data array. For the conditions tested, we found no clear advantage of this method over the current ISO procedure. For other applications, the slanted-edge analysis provides a front end to Tatian's method, and others, based on parametric modeling or statistical fitting.