A comparison of linear and non-linear magnification on target detection performance in visual inspection

Considering the importance of magnification on visual inspection, two experiments on the detection of peripheral visual targets were conducted in this study: one on non-linear and the other on linear magnification. In the first experiment, two levels of non-linear scaling were compared with non-scaling. Performance with scaled stimuli was better than that with non-scaled stimuli, on each of the four axes tested. An unexpected decline in performance was found at the least eccentric target location for the highest level of scaling. There was a performance increment at the most eccentric target locations for both experiment investigated the effect of four levels of linear magnification and found that performance with scaled stimuli was better than performance with non-scaled stimuli. However, linear scaling of objects at central locations to the same degree as those in the periphery yielded little benefit here. At greater eccentricities, the effect of magnification of objects appeared to be offset to some extent by lateral interference. Inter-object spacing was magnified to the same extent as objects, so that the number of object locations decreased as scaling increased and the numbers of correctly detected target locations were similar for all scaling conditions. Both linear and non-linear magnification facilitated peripheral target detection performance but, perhaps because of the effects of lateral masking, neither linear nor non-linear magnifications seems to nvolve a trade-off between the number of objects in the viewing area and the scaling level to attain acceptable performance with least magnification. However, it seems wasteful to enlarge targets at the vicinity of fixation to the same degree as those at more eccentric locations. Appropriate inter-object spacing is important as well as object size when considering magnification conditions to optimize inspection performance.