Artificial intelligence-based unmanned aerial vehicle engine blade defect rapid and accurate identification method

By using an artificial intelligence-based method to extract sub-pixel edge features of UAV engine blades and combine them with centrifugal force fields, the problem of dynamic assessment of blade defect identification in existing technologies has been solved, enabling accurate identification and risk assessment of blade defects.

CN122199490APending Publication Date: 2026-06-12HEBEI RUILAI AVIATION EQUIP CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
HEBEI RUILAI AVIATION EQUIP CO LTD
Filing Date
2026-03-16
Publication Date
2026-06-12

AI Technical Summary

⚠Technical Problem

Existing technologies for identifying blade defects under high-speed rotation conditions rely on static image size measurements, which are insufficient to reflect the expansion trend of defects under alternating loads. This leads to the missed or misjudged detection of minute initial cracks, resulting in mechanical failures.

⚗Method used

An artificial intelligence-based approach is adopted to extract the pixel matrices of the leading and trailing edges of UAV engine blades, perform gradient operator convolution and morphological closing operations to generate a set of sub-pixel edge coordinate points, establish an analytical continuous curve by combining cubic spline function operations, calculate the local curvature value, and generate a crack propagation risk index by combining the centrifugal force field vector to achieve dynamic assessment.

🎯Benefits of technology

It improves the reliability of blade defect identification, eliminates the limitations of relying solely on static visual dimensions, and enables accurate identification of dynamic failure risks.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122199490A_ABST
    Figure CN122199490A_ABST
Patent Text Reader

Abstract

The present application relates to the technical field of image recognition, in particular to a method for quickly and accurately identifying defects of a UAV engine blade based on artificial intelligence, comprising the following steps: extracting a pixel matrix of the leading edge and trailing edge of the UAV engine blade, performing gradient operator convolution and morphological closing operation, and generating a rough edge pixel set of the blade; calling a straight line segment geometric vector and a straight line segment pixel length value to calculate a radial component to generate a centrifugal force field vector, fuse the sine value of the included angle and the leaf root distance weight to accumulate and calculate a crack propagation risk index, compare it with a set fatal threshold to generate a blade defect rating classification parameter, couple the geometric features of the static image with the centrifugal force field distribution under the dynamic rotating working condition of the engine, establish a stress state driven risk assessment mechanism, eliminate the limitations of relying on single static visual size for judgment, and improve the reliability of dynamic failure risk identification.
Need to check novelty before this filing date? Find Prior Art