Defect detection of reactor pressure vessel sealing surface based on deep learning

Detecting the sealing surface of the reactor pressure vessel is essential for the safety and protection of nuclear energy. The traditional manual detection is inefficient and may cause irreparable harm to the human body. To solve these problems, this paper proposes an inspection method based on machine vision to replace manual inspection. This paper designs and compares deep learning algorithms that use different parameters and different processing functions for defect detection, and use Python language to complete data preprocessing, model training, and model prediction. Finally, the feature extraction and classification of the collected image are realized, and the defect detection of the sealing surface is completed.