Portunus trituberculatus, a crustacean species of significant economic value, is important in aquaculture and seafood processing. Accurate gender classification is critical for optimizing aquaculture strategies and improving operational efficiency in these industries. Traditional manual methods for sex classification are labor-intensive and prone to errors, highlighting the need for automated solutions. This study proposes an automated gender classification method based on deep learning, utilizing the SE-ResNet18 model, an enhanced variant of ResNet-18. The SE-ResNet18 model incorporates the Squeeze-and-Excitation (SE) module and global average pooling, enabling it to selectively emphasize key feature channels. The model was trained and validated on a large dataset of male and female Portunus trituberculatus images, with data augmentation techniques applied to improve generalization. The results show that SE-ResNet18 achieves a classification accuracy of 99.5%, nearly 4 percentage points higher than ResNet-18. Specifically, male crabs were classified with 99.68% accuracy, and female crabs with 99.74%. The model's robustness was tested under varying conditions, confirming its suitability for real-world applications in automated seafood processing and aquaculture. In conclusion, SE-ResNet18 offers a highly accurate and scalable solution for gender classification in Portunus trituberculatus, with the potential to significantly enhance productivity and efficiency in the aquaculture industry.