Online Color Classification System of Solid Wood Flooring Based on Characteristic Features

Area of Science:

• Computer Vision
• Machine Learning
• Materials Science

Background:

• Solid wood flooring requires color sorting for aesthetic interior design.
• Manual sorting is costly and inefficient, hindering customization.
• Automating this process is crucial for meeting diverse customer needs.

Discussion:

• Machine vision and deep learning methods, including VGG16, DenseNet121, and XGBoost, were employed for automated color classification.
• Data augmentation techniques were used to expand the dataset of 108 wood floor images.
• The study compared the performance of different deep learning models for color grading.

Key Insights:

• XGBoost demonstrated high accuracy (97.22%) in classifying solid wood flooring colors.
• The XGBoost model achieved rapid processing with an average test time of 51 ms per image.
• The developed system significantly reduces manual sorting costs and enhances production efficiency.

Outlook:

• This automated approach can be applied to other wood-based materials requiring color grading.
• Further research could explore real-time sorting integration on production lines.
• Optimizing deep learning models may lead to even higher classification accuracy and efficiency.