Gamma and X-rays are being widely used in industrial applications for non-imposing imaging as part of non-intrusive testing. However, clarifications about X-ray images remain subjective because it heavily relies on individual competence. Therefore, to interpret X-ray images, several machine learning algorithms offer a solution by utilizing collective learning to enhance the speed and accuracy of identifying defects, classification, grading, and more. Various image analysis models developed over the years have been reviewed in this paper, hence, providing a comparative analysis based on different parameters such as their pros, cons, and the suitability for specific applications. We aim to enhance the speed and accuracy of detecting the anomalies like round, inclusions, undercut, concavity, burn through and so on, through automation, thereby reducing the instances of errors during manual detection which could further aid in multiple industrial applications, mainly ship building. By using machine learning and deep learning algorithms, this research aims to significantly improve quality assurance procedures in industrial applications.