Exploring Visual Aesthetics in Ceramic Packaging Through Computational Style Analysis

Abstract

The packaging is defined as ceramic packaging because ceramic materials are incorporated into several structural components. Innovative product packaging plays an essential role in driving commercial value by increasing product sales and strengthening brand recognition. Compared with plastic packaging, ceramic packaging faces greater challenges due to its heavier weight and higher production and testing costs. To address these challenges, this study proposes a Spiral-Optimized Adjustable XGBoost (SO-AXGBoost) model to evaluate ceramic packaging design. A large dataset of images containing diverse ceramic packaging styles was collected, and feature extraction was carried out using Discrete Wavelet Transform (DWT). The experimental results indicate that the optimized model significantly improves the accuracy and effectiveness of ceramic packaging evaluation compared to the baseline method. The findings verify that integrating advanced learning techniques with visual communication style analysis is feasible and beneficial. This novel approach enhances not only the functional and aesthetic performance of ceramic packaging but also contributes to improved market competitiveness and alignment with consumer preferences