Due to the recent increase in traffic accidents, pelvic fractures are increasing, second only to skull fractures in terms of mortality and complication risk. Research is also being actively conducted on the treatment of intra-abdominal bleeding, which is the primary cause of death related to pelvic fractures. A considerable amount of preliminary research has been conducted on segmenting tumors and organs. However, studies on clinically useful algorithms for bone and pelvic segmentation based on the developed models were limited. In this study, we explored the potential of deep learning models based on previous studies to achieve accurate pelvic region segmentation in X-ray images. Data were collected from X-ray images of 940 patients aged 18 or older at Gachon University Gil Hospital from January 2015 to December 2022. To segment the pelvis, Attention U-Net and Swin U-Net were learned and compared and analyzed through 5-Fold cross-validation. It was found that the Swin U-Net model had relatively high performance compared to Attention U-Net. The Swin U-Net model achieved an average sensitivity of 96.77%, an average specificity of 98.50%, an average of 98.03%, and an average DSC of 96.32%.