This　paper　presents　a　geomagnetic　detection　method　for　pipeline　defects　using　complete　ensemble　empirical　mode　decomposition　with　adaptive　noise　(CEEMDAN)　and　wavelet　energy　product　(WEP)　-　Teager　energy　operator　(TEO),　which　improves　detection　accuracy　and　defect　identification　ability　as　encountering　strong　inference　noise.　The　measured　signal　is　first　subtly　decomposed　via　CEEMDAN　into　a　series　of　intrinsic　mode　functions　(IMFs),　which　are　then　distinguished　by　the　Hurst　exponent　to　reconstruct　the　filtered　signal.　Subsequently,　the　scale　signals　are　obtained　by　using　gradient　calculation　and　discrete　wavelet　transform　and　are　then　fused　by　using　WEP.　Finally,　TEO　is　implemented　to　enhance　defect　signal　amplitude,　completing　geomagnetic　detection　of　pipeline　defects.　The　simulation　results　created　by　magnetic　dipole　in　a　noisy　environment,　indoor　experiment　results　and　field　testing　results　certify　that　the　proposed　method　outperforms　ensemble　empirical　mode　decomposition　(EEMD)-gradient,　EEMD-WEP-TEO,　CEEMDAN-gradient　in　terms　of　detection　deviation,　peak　side-lobe　ratio　(PSLR)　and　integrated　side-lobe　ratio　(ISLR).