基于SSD算法的航空发动机内部凸台缺陷检测

Defect detection of aircraft engine internal convex based on SSD algorithms

基于深度学习的背景,提出将目标检测算法用于航空发动机内部凸台缺陷的检测研究。首先介绍了算法的主要特点,通过使用聚类分析方法改进算法产生默认框的生成方式,提高了算法模型对发动机内部凸台缺陷的匹配能力;并采用多种图像处理算法相结合,对目标图像进行预处理来突出凸台缺陷的主要特征,增强了算法模型提取待检测目标的特征信息,从而进一步提高检测算法对于航空发动机凸台缺陷的检测精度。最终检测算法对于凸台缺陷的检测精度达到了95%以上。

Based on the background of deep learning,the target detection algorithm is proposed to detect the internal boss defects of aeroengine.Firstly,the main characteristics of the algorithm are introduced.By using clustering analysis method to improve the method of generating default frame,the matching ability of the algorithm model to engine internal boss defects is improved;and a variety of image processing algorithms are combined to preprocess the target image to highlight the main features of boss defects,which enhances the algorithm model to extract the features of the target to be detected Information,so as to further improve the detection accuracy of detection algorithm for Aeroengine boss defects.The detection accuracy of the final detection algorithm for convex defects is over 95%.