Terahertz　(THz)　imaging　has　been　widely　used　in　industrial　nondestructive　testing　(NDT)　of　nonpolar　materials　due　to　its　unique　property.　Minor　defect　detection　via　THz　NDT　at　high　accuracy　and　fast　speed　is　essential　for　industrial　online　detection　systems.　However,　traditional　defect　detection　algorithms　cannot　meet　the　demand　of　real-time　high-precision　detection　of　minor　defects.　Therefore,　based　on　the　you　only　look　once　x　(YOLOX)　algorithm　and　multiscale　attention　(MSA)　mechanism,　the　modified　YOLOX　network　called　YOLOX-MSA　is　proposed　as　a　one-stage　minor　defect　detection　framework　to　improve　the　detection　accuracy　while　supporting　the　real-time　operation.　The　proposed　YOLOX-MSA　network　improves　the　mean　average　precision　(mAP)　by　at　least　11.65%　on　the　printed　circuit　board　(PCB)　dataset　with　THz　characteristics　when　the　intersection　over　union　(IoU)　is　0.75.　In　addition,　the　proposed　algorithm　can　reach　the　detection　speed　as　24-25　frames　per　second　(FPS).　Overall,　our　proposed　method　can　be　beneficial　to　generalize　the　THz　NDT　in　the　frequency　domain　on　the　minor　defects　of　nonpolar　material,　which　will　fulfill　the　impending　requirements　of　real-time　defect　detection　for　industrial　applications.