Scanpath Theory, Attention, and Image Processing Algorithms for Predicting Human Eye Fixations

Attention has two strong links to the top-down scanpath theory. The initial intentional conscious organization of the perceptual cognitive-spatial model that directs vision must occur at a high level of cortical processing. Then, the directions of the scanpath eye movements controlling successive attentional foveations toward the informative regions-of-interest enable matching and checking of high-resolution detail from bottom-up sensory signals. The scanpath theory firstly takes into account the dichotomy between low-resolution peripheral vision and high-resolution foveal vision; this dualism necessitates an important role for eye movements. It further develops that representation over distributed modules of the cortex that generate the complex model of perception as an active process. Finally, the scanpath controls eye movement foveations that are the usual and natural means of directing visual attention. Understanding the role of the attention shifts–eye movements' scanpath in human vision is an important step toward the achieving of independent and automatic image processing in the computer vision community. It has been demonstrated that a small and manageable collection of image processing algorithms, experimentally selected and then combined together can serve in a task such as predicting human eye fixations.