A deep learning model for cigarette package surface defect detection integrating multi-scale features

The detection of surface defects on cigarette packages plays a crucial role in ensuring product quality and maintaining manufacturing efficiency. Traditional methods often struggle with variations in defect size, shape, and texture, leading to inconsistent detection performance. This paper presents a deep learning model that integrates multi-scale feature extraction to address these challenges. The proposed model combines convolutional neural networks (CNNs) with multi-scale feature fusion techniques to enhance its capability in capturing both fine and coarse defect patterns. By employing a multi-branch architecture, the model effectively processes features at different scales, improving the detection of diverse defect types. Experimental results demonstrate that the proposed approach achieves superior performance compared to conventional models, with higher precision, recall, and F1-score across multiple datasets. This research contributes a robust and scalable solution for surface defect detection, with potential applications in automated quality control systems.