Abstract
Occluders are prevalent in the visual environment. For example, we need to see through the rain drops on the windshield when driving, or see through the blinds from our living room. Some occluders are random (e.g., rain drops), while others are structured (e.g., blinds). How does occluder pattern influence visual perception? To answer this question, we presented occluded scenes to participants and asked them to judge whether the scene was indoors or outdoors in Experiment 1. The pattern of the occluders was either random or structured (e.g., stripy). The occluders were either chunky (leaving large chunks of the scene visible), or fine-grained (similar to mesh). Scene discrimination was better when the occluders were structured than random but only for fine-grained occluders, and it was identical for structured and random occluders when they were chunky. This suggests that fine-grained structured occluders are more visually penetrable than random occluders. For chunky occluders, the need for visual penetration is reduced due to the large intact scene chunks, which could explain the lack of difference. To see whether this effect is specific to scene perception, we replicated the experiment using object images where participants viewed occluded objects and judged whether the object was animate or inanimate in Experiment 2. The results were highly similar: object discrimination was better when fine-grained occluders were structured than random, but performance was comparable for chunky structured and random occluders. Finally, we extended our findings to numerosity perception where participants estimated the number of objects in an occluded array in Experiment 3. Number estimation was more accurate when fine-grained occluders were structured than random. Therefore, for all three types of perception (scene, object, number) fine-grained structured occluders are more visually penetrable than random occluders. The findings reveal new insights into how visual perception operates with incomplete information.