This paper proposes an efficient method for detecting defects in the outer packaging of mobile phones. The Faster RCNN algorithm is utilized as the primary detection method, which trains a deep neural network to automatically detect defects in mobile phone outer boxes. Compared to the traditional manual detection method, the deep learning-based approach exhibits stronger adaptive abilities and robustness. Our method employs the resnet-50 network with residual units as the backbone extraction network, and utilizes the cross-entropy loss function to train the classifier. This approach effectively addresses the diversity and complexity issues associated with the outer packaging box of mobile phones, resulting in improved detection performance and efficiency. This paper presents a comparative analysis and experimental verification of a proposed method for defect detection in the outer packaging box of mobile phones. The results demonstrate the feasibility and effectiveness of the proposed method, which accurately detects various defects and shows excellent performance in terms of detection speed, accuracy, and stability. The study highlights the superiority of the proposed method in the field of defect detection in the outer packaging box of mobile phones.