A Novel Fractional-Order Chaotic Phase Synchronization Model for Visual Selection and Shifting

Area of Science:

• Cognitive Informatics
• Computational Neuroscience
• Image Processing

Background:

• Visual information processing is a key area within cognitive informatics.
• Existing object selection models lack mechanisms for simulating visual shifting.
• Understanding the human visual system's cognitive mechanisms is crucial.

Purpose of the Study:

• To establish a two-layer fractional-order chaotic network model simulating visual selection and shifting.
• To introduce control units for object selection within the model.
• To provide a more biologically plausible model of human visual information processing.

Main Methods:

• Developed a two-layer fractional-order chaotic network.
• Implemented image segmentation in the first layer using a chaotic network.
• Introduced a control layer with a central neuron for object selection and shifting.
• Proposed a synchronization strategy between subnets and the central neuron.

Main Results:

• The model successfully simulates the mechanism of visual selection and shifting.
• The proposed control units effectively manage object selection.
• Experimental verification using artificial and natural images confirmed the model's reasonability.
• The model demonstrates better correspondence with the human visual system than traditional models.

Conclusions:

• The developed chaotic network model offers a novel approach to simulating human visual cognition.
• The model provides new possibilities for analyzing the human cognitive system's mechanisms.
• This research enhances our understanding of visual information processing and selection.