Exploration of Visual Attention Analysis and Optimization Algorithms in Packaging Design

In today’s highly competitive market, consumer decisions are increasingly shaped by evolving preferences and the rapid recognition of brands. To meet these demands, this study proposed an innovative approach that integrates visual attention analysis with advanced optimization algorithms to enhance packaging efficiency and consumer engagement. A comprehensive packaging visual attention and engagement dataset was developed, encompassing design attributes, attention metrics, emotional responses, engagement measures, and performance indicators. During preprocessing, all package images were resized and normalized to ensure uniformity. Convolutional Neural Networks (CNNs) were employed for feature extraction, enabling the identification of critical design elements such as layout, color palette, typography, and branding components.To further improve design optimization, the U-Net Driven Multi-Objective Cuckoo Search Tuned Efficient Fire Hawk Optimizer (UN-MOCS-EFHO) approach was implemented. This method combined the strengths of U-Net for segmentation, MOCS for multi-objective optimization, and EFHO for efficient convergence. Visual saliency was defined as the ability of specific design elements-such as color, contrast, or layout harmony-to capture consumer attention, while emotional resonance referred to how design structure influenced consumer perception and brand recall. By jointly modeling these factors, the proposed method optimized layout parameters and aesthetic features to achieve maximum consumer appeal.Experimental results demonstrated that the UN-MOCS-EFHO model outperformed other approaches, achieving superior performance with 99.995% accuracy, 98.5% precision, 97.4% recall, and 98.8% F1-score. This research established a data-driven, intelligent design framework that leverages algorithmic modeling to create personalized, visually engaging, and emotionally resonant packaging aligned with contemporary market trends.