Deep learning defect automatic detection and identification method based on small sample aero-engine blade CT image

Abstract

The invention discloses a deep learning defect automatic detection and identification method based on a small sample aero-engine blade CT image. The method comprises the following steps of carrying out digital processing on a blade CT film, manually calibrating the type and position of each defect to establish a defect sample label set, cutting a local defect area image of the blade, and performing data expansion and corresponding label correction expansion to establish a deep learning model training sample set, constructing a deep learning aero-engine blade defect detection and identification network, training a deep learning aero-engine blade defect detection and identification network, establishing an automatic detection and identification model according to the defect detection and identification network and the final training parameters, and inputting the CT image into the defect detection and identification model to automatically detect, identify and position the blade defect. According to the method, the problem that the number of defective blade samples is small is solved, the influence of human factors is overcome, and the radiographic detection efficiency of the aero-engine blade and the detection precision of tiny defects are greatly improved.

Description

technical field [0001] The invention belongs to the field of aero-engine blade processing, manufacturing and quality inspection, and in particular relates to a deep learning defect automatic detection and recognition method based on CT images of small-sample aero-engine blades. Background technique [0002] Aeroengine blades are the main load-bearing components in the working process of the engine, and the quality of the blades is closely related to the safe operation of the aeroengine. The current non-destructive testing techniques (ray testing, eddy current testing, magnetic particle testing, penetration testing, etc.) are widely used in the field of aero-engine blade testing. With the continuous development of the aviation industry, while the performance of aircraft continues to improve, higher requirements are placed on the reliability of aeroengine blades, and the quality inspection requirements are also becoming more stringent. [0003] Restricted by the casting proce...

AI Technical Summary

Problems solved by technology

The missed detection rate and false detection rate of these traditional defect detection methods are still high, and cannot effectively improve the quality of turbine blades

[0004] With the advent of the era of artificial intelligence, deep learning technology has been gradually applied to defect detection, and has been used in the identification of defects with obvious morphological characteristics such as welds and holes. However, for most defect detection, the traditional non-destructive testing technology

The casting process of turbine blades is complicated, and there are many types of typical defects. For a long time, the defect detection of turbine blades has been evaluated by manual experience

Images