Wire harness production defect detection method

By combining a global shutter CMOS camera and an improved YOLOv7 model with stroboscopic fill light technology, wavelet transform, and time-frequency cross-correlation analysis, the problems of insufficient detection speed and accuracy at the wire harness production line end are solved, and efficient, real-time defect detection of wire harnesses is achieved.

CN120741513AInactive Publication Date: 2025-10-03深圳市揽英科技有限公司
1 Cites 1 Cited by

Patent Information

Application Number
CN202511135040.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-08-14
Publication Date
2025-10-03
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

The existing technology lacks detection speed and accuracy at the wire harness production line end, and cannot meet the immediacy and accuracy requirements of intelligent production lines for quality control, especially when detecting defects such as terminal crimping misalignment, outer insulation layer damage, and abnormal internal circuit conductivity.

Method used

A global shutter CMOS camera and an improved YOLOv7 model are used for external image detection, and stroboscopic fill light technology is used to obtain the initial wire harness image. The incident test wave signal is used for frequency scanning, and the cable reflection signal is processed using wavelet transform and time-frequency cross-correlation analysis. An adaptive longicorn whisker search algorithm is used to accurately locate the fault position and filter out false signals.

Benefits of technology

It realizes high-speed, real-time detection of the wire harness production line end, improves the detection accuracy of defects such as abnormal conductivity of the external insulation layer and internal line, and meets the quality control needs of the intelligent production line.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention discloses a wire harness production defect detection method, and relates to the technical field of cable detection, and the method comprises the steps: S1, deploying a high-speed camera at a production line terminal, and collecting an initial wire harness image; s2, analyzing the initial wire harness image through an improved YOLOv7 model to obtain an external detection result; s3, performing frequency sweeping on the wire harness through the incident test wave signal to obtain a wire harness time domain waveform, and obtaining a cable reflection signal; s4, performing first synchronous compression wavelet transform processing on the cable reflection signal to obtain a first reconstruction time-frequency spectrum; performing first time-frequency cross-correlation analysis on the first reconstruction time-frequency spectrum to obtain a primary screening fault position set; screening all point positions in the primary screening fault position set through a self-adaptive beetle antennae search algorithm to obtain filtered reflection signals; and performing third synchronous compression wavelet transform processing and second time-frequency cross-correlation analysis on the filtered reflection signal to obtain a detection fault position set. Detection of internal and external defects of the wire harness is efficiently and accurately completed.
Need to check novelty before this filing date? Find Prior Art