Visual Categorization with Aerial Photographs

This article investigates how people process information from aerial photographs to categorize locations. Three cognitive experiments were conducted with human subjects viewing a series of aerial photographs and categorizing the land use for target locations. Reaction time, accuracy, and confidence were considered as dependent variables related to the success of the categorization process. The first experiment considered two categories of land use, the relative size of the visual field, and two rounds of unsupervised learning. Subjects were more successful categorizing higher-order land-use classes than they were lower-order categories. Subjects were significantly more accurate and confident with larger photographs, but not significantly faster. Significant improvements between the rounds indicated unsupervised learning was taking place. The second experiment confirmed the hypothesis that geographers would have more success than nongeographers during a single categorization round. The types of land use considered were significantly related to success. Subjects were again more accurate and confident, but not faster with larger visual fields. A third experiment considered seven rounds of supervised learning, the sex of the subjects, and the amount of experience with photographs. Reaction time and confidence improved with supervised learning, but accuracy did not. Subjects had significantly more success with photographs they viewed more than one time. Male subjects were significantly faster, more accurate, and more confident than female subjects at doing the categorization task. By the seventh learning round the male advantage in reaction time and accuracy was no longer significant, but the male advantage in confidence continued through seven learning rounds.