The following is a summary of “Texture statistics involved in specular highlight exclusion for object lightness perception,” published in the March 2023 issue of Ophthalmology by Nohira, et al.
The human visual system has the remarkable ability to estimate the physical properties of objects, such as their lightness. Previous studies have shown that when it comes to glossy three-dimensional objects, the specular highlights are detected and excluded in lightness perception. However, the mechanisms underlying this exclusion have not been fully understood, and this is what the current study aimed to investigate.
To do this, the researchers used Portilla-Simoncelli texture statistics (PS statistics), which is a set of image features that closely resembles the representation in the early visual cortex. They explored the relationships between highlight exclusion and four subsets of image features using two experiments.
In the first experiment, computer-generated images of plastic plates with different physical parameters were used as stimuli, and a lightness-matching task was used to measure lightness perception. The researchers then calculated the highlight exclusion index, which represented the degree of highlight exclusion, and evaluated the correlation between this index and the four subsets of PS statistics. In the second experiment, synthesized images were created using an image synthesis algorithm, and the highlight exclusion index was measured after manipulating each of the four PS statistic subsets.
The results of the study revealed that the subset of PS statistics consisting of lowest-order image features, such as moment statistics of luminance, was the most important factor for highlight exclusion. The other three subsets consisting of higher-order features were not as crucial. These findings suggested that low-order image features are necessary for highlight exclusion, although higher-order image features may also be involved in the process.
In conclusion, the study provided insights into the mechanisms underlying the exclusion of specular highlights in lightness perception of glossy 3D objects. The results suggested that low-order image features play a crucial role in this process, which may have implications for the development of computational models of visual perception.