As a key element in fashion design, the precise identification of suit collar shape is of great significance for the intelligent development of the fashion industry. This study selected Faster R-CNN as the main model to construct the Blaze6480 image dataset, which includes four types of collar types: flat collar, ridge collar, green fruit collar, and no collar. At the same time, accuracy (A), precision (P), recall (R), average precision (AP), mean accuracy (mAP), and F1 score key indicators were used to comprehensively evaluate the performance of the Faster R-CNN model in women’s suit collar recognition tasks. The experimental results show that Faster R-CNN performs well in both recall rate (73.8%) and average accuracy (mAP 73.5%) in complex background collar detection, but the accuracy (62.1%) still needs to be optimized. This study provides technical reference for clothing attribute recognition and practical significance for automated classification of collar shapes in scenarios such as e-commerce platforms and intelligent clothing manufacturing.