Pipeline defect detection method and system based on multi-source data fusion

The invention discloses a pipeline defect detection method and system based on multi-source data fusion, and belongs to the technical field of image processing. Comprising the following steps: acquiring a visible light image and an infrared image of a to-be-detected pipeline, and processing the visible light image and the infrared image through a self-adaptive channel addition fusion strategy to generate a pipeline defect detection image; sequentially processing the pipeline defect detection images through a plurality of cascaded CBM modules and CBMC3 modules to obtain pipeline defect feature maps of different scales; inputting the pipeline defect feature maps of different scales into a feature fusion network, and performing multi-layer feature fusion on the pipeline defect feature maps from bottom to top to obtain pipeline defect feature fusion maps of different scales; and obtaining a pipeline defect detection result according to the pipeline defect feature fusion images of different scales. Defect features