The Human Visual System and the Perception of 3D Objects

               The ability of the human visual system to perceive 3D shape is fascinating. It is safe to say that perception plays an essential role in our everyday lives, at every moment. In fact, the visual system is at work even now, granting you the ability to perceive the laptop or smartphone from which you may be reading this blog. What is most interesting about perception however, are the processes underlying it. For years on end, researchers have been carefully studying the intricacy of the visual system and its ability to interpret shape. Several conclusions have been reached as to how this may be possible, yet as humans, we cannot help but wonder: in what way do each of these processes play a specific role in our perception? And how can this be used to help our well-being?

                In the paper "The human visual system's assumption that light comes from above is weak", researchers Yaniv Morgenstern et al. studied the ability of the human visual system to perceive 3D shape using the light-from-above prior assumption and lighting cues. Surprisingly, their results indicated that the preconceived notion of the light-from-above prior playing a critical role in perception was weak. On the contrary, Morgenstern et al. found that the human visual system has a much heavier reliance on lighting cues when determining shape. The influence of lighting cues was so profound, that researchers discovered the weakest lighting cues had a similar effect as the light-from-above prior on perception. They were able to reach that conclusion following a series of shape-from-shading conditions, while simultaneously measuring the angles of effective lighting direction relative to the prior. Under the weak lighting cue condition, researchers observed a 66º circular standard deviation from the width of the light-from-above prior. A comparison was then drawn between this condition and the circular SD of lighting directions that are observed in our everyday lives, which was stated as 54º. The comparatively small variation in the angles of effective lighting direction in relation to the prior under weak cues and those apart of everyday life were therefore, key in establishing the strong influence lighting cues have on perception. 

                Similar to the work by Morgenstern et al., an article reviewed by Emily Henderson titled "Humans may perceive thickness based on image characteristics" for News Medical Life Sciences detailed the manner in which researchers Ohara, Koida, and Kim went about tackling the topic of perception. Interestingly, they approached from a completely different perspective and illustrate yet another manner in which the visual system works to perceive characteristics of 3D shape. In their research, they studied the effects of image cues (such as transparent properties) on 3D shape perception--specifically, a judgment of how thick an object appears to be under different conditions. Employing a computer driven experimental method, Ohara et al. discovered that participants displayed a higher tendency to judge transparent objects as flatter, in comparison to those with a glossy appearance. Furthermore, following an analysis of these images, researchers were then able to determine that the most precise estimations of thickness were linked to regional variation. In short, similar to how Morgenstern et al. observed angles of lighting direction relative to the prior for weak cue conditions, Ohara et al. found that regional variation from local contrast played a key role in participants' judgments. Although it was mentioned in the article that a clear cause for misinterpretation is not yet known, it is interesting to consider the fact that the thickness of transparent objects may have been more likely to be misjudged due to the lower local contrast caused by this transparency. This is supported by the lack of pixel contrast in a confined part of an image that displays a transparent object. The lack of pixel contrast makes it harder for the participant to discriminate shape and therefore, misinterpret thickness. 

                Regardless, both articles successfully illustrated different manners in which the human visual system is able to perceive 3D shape. As mentioned in the Henderson article, further understanding of how misinterpretations in shape may arise is crucial in developing technology that will aid those who are visually impaired. For example, Henderson mentions the use of a computational model in the form of glasses, or even mobility devices. However, the different approaches to perception in both articles provide two avenues for aid. Not only would technology-driven glasses serve the purpose of helping those who may not be able to discriminate between bumps and flat areas of a surface, but the incorporation of a tint to the glasses may help those who suffer from not being able to determine the shape of an object as a result of insensitivity to lighting cues. Our ability to provide variations of aid therefore, reinforces the intricacy of the visual system when it comes to 3D shape perception. 

                                                                            References 

Morgenstern, Y., Murray, R.F., Harris, L.R. (2011). The human visual system's assumption that light comes from above is weak. Proceedings of the National Academy of Sciences, 108(30), 12551-12553; DOI:10.1073/pnas.1100794108

Henderson, E. "Humans may perceive thickness based on image characteristics". News Medical Life Sciences, 11 February 2021, https://www.news-medical.net/news/20210211/Humans-may-perceive-thickness-based-on-image-characteristics.aspx