The invention discloses a strain clamp defect unmanned aerial vehicle detection method fused with multi-modal data, which comprises the following steps: establishing a data set which comprises an X-ray image, an ultrasonic image and a hydraulic strain clamp image, and constructing a strain clamp defect detection model fused with the multi-modal data; inputting the data set into a strain clamp defect detection model fused with the multi-modal data for training to obtain a trained strain clamp defect detection model fused with the multi-modal data, inputting the X-ray image and the ultrasonic image into the trained strain clamp defect detection model fused with the multi-modal data at the same time for processing to obtain a Grad-CAM heat map, and obtaining the strain clamp defect detection model fused with the multi-modal data. And processing the Grad-CAM heat map into the size of a hydraulic strain clamp map, carrying out fusion imaging, and displaying and outputting a detected defect type and a judgment position on the fusion imaging. According to the invention, intelligent defect detection of the hydraulic strain clamp is realized by fusing the ultrasonic image and the X-ray image.

    本发明公开了一种融合多模态数据的耐张线夹缺陷无人机检测方法,包括以下步骤,建立数据集,数据集包含X光图像、超声图像和液压型耐张线夹图,构建融合多模态数据的耐张线夹缺陷检测模型;将数据集输入融合多模态数据的耐张线夹缺陷检测模型进行训练,获得训练后的融合多模态数据的耐张线夹缺陷检测模型,将X光图像和超声图像同时输入到训练后的融合多模态数据的耐张线夹缺陷检测模型中进行处理,获得Grad‑CAM热图,将Grad‑CAM热图处理为液压型耐张线夹图的大小并融合成像,在融合成像上显示输出检测的缺陷类型及判断位置。本发明通过融合超声图像和X光图像实现液压型耐张线夹缺陷智能检测。


    Zugriff

    Download


    Exportieren, teilen und zitieren



    Titel :

    Multi-modal data fused strain clamp defect unmanned aerial vehicle detection method


    Weitere Titelangaben:

    一种融合多模态数据的耐张线夹缺陷无人机检测方法


    Beteiligte:
    WU SHIYOU (Autor:in) / LIU ZHOU (Autor:in) / ZHOU FUWEN (Autor:in) / SHANG WENHUI (Autor:in) / FU LIXIANG (Autor:in) / LI FAN (Autor:in) / LI MING (Autor:in) / ZHANG ZHENG (Autor:in) / ZENG CHUN (Autor:in) / FU SHIXUN (Autor:in)

    Erscheinungsdatum :

    25.04.2025


    Medientyp :

    Patent


    Format :

    Elektronische Ressource


    Sprache :

    Chinesisch


    Klassifikation :

    IPC:    G06V / B64U / G06N COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS , Rechnersysteme, basierend auf spezifischen Rechenmodellen / G01N Untersuchen oder Analysieren von Stoffen durch Bestimmen ihrer chemischen oder physikalischen Eigenschaften , INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES



    FPV auxiliary mounting system for strain clamp unmanned aerial vehicle ray detection

    YUE XIANQIANG / WANG CHI / ZHANG HUA et al. | Europäisches Patentamt | 2022

    Freier Zugriff


    Unmanned aerial vehicle propeller clamp device

    HUANG HAIJUN / WU XIAOGUANG / WU MIN et al. | Europäisches Patentamt | 2024

    Freier Zugriff

    Pipeline defect detection method based on unmanned aerial vehicle spectrum and unmanned aerial vehicle

    WANG YANFEI / LIU LIBO / HU QIANG | Europäisches Patentamt | 2023

    Freier Zugriff

    Unmanned aerial vehicle for wire clamp X-ray detection

    WANG WEI / XU XIAOBIN / ZHAO XI et al. | Europäisches Patentamt | 2021

    Freier Zugriff