Machine Vision-Based Tenon Size Detection for Aero-Engine Blades Using Enhanced Edge Detection and Fitting Algorithms

Abstract

Accurate dimension detection of the tenon in aero-engine blades is critical to ensuring proper fitting and performance. Traditional methods, such as the enlarged projection technique, are becoming inadequate as the demand for qualified blades grows. This paper presents a machine vision detection system designed to address the limitations of current tenon size detection methods. The system employs an improved bilateral filtering adaptive Canny operator for edge detection and combines Hough transform with least squares for edge fitting. The proposed algorithm enhances the accuracy of tenon dimension detection by retaining more edge feature information and adjusting the threshold adaptively to match image features. Experimental results show that the method achieves higher precision and execution efficiency compared to traditional techniques, providing more reliable contour detection results. Despite its success, further improvements in image acquisition quality and edge extraction accuracy will be pursued to meet the increasing demand for precision in future applications.