Developments in Digital Image Processing (DIP) have become pivotal across various domains, including medical imaging, remote sensing, industrial automation, and computer vision. To enhance decision-making in specific fields, DIP aims to segment images into meaningful regions or objects for interpretation. In the context of weld radiography, image segmentation is crucial for spotting weld flaws, which can significantly impact the quality and safety of the manufacturing process. Despite the availability of many segmentation techniques, deciding the most suitable method for a specific domain is critical to achieve the highest accuracy. This selection process necessitates a thorough understanding of both the segmentation methods and the unique characteristics of the domain. Specifically, in non-destructive testing (NDT), the segmentation process plays a significant role on spotting weld flaws from industrial X-rays thereby ensuring the structural integrity of materials used in manufacturing. This study evaluates the efficacies of different image segmentation techniques in identifying defects from weld radiographic images. The findings of this study suggest segmentation methods tailored to the specific characteristics of weld defects, which enhances both segmentation accuracy and defect classification in weld radiographs. To evaluate the process carried out, the segmentation process of image analysis is assessed using Root Mean Square Error (RMSE), cumulative distribution, and threshold difference. Additionally, the F1 score is computed to assess classification accuracy and determine the most suitable segmentation methods for industrial radiography applications.