In　view　of　the　problem　that　magnetic　acoustic　emission　(MAE)　signal　has　strong　background　noise　and　characteristic　parameters　cannot　be　effectively　extracted　from　the　signal,　the　signal　was　decomposed　by　complementary　empirical　modal　decomposition　(CEEMD).　The　decomposed　IMF　components　were　analyzed　with　the　original　signal,　and　the　IMF　components　with　large　correlation　coefficients　to　the　original　signal　were　retained　and　added　after　the　characteristic　parameters　are　calculated.　Therefore,　the　purpose　of　effectively　suppressing　noise　can　be　achieved.　A　method　for　extracting　MAE　characteristic　parameters　based　on　CEEMD　was　proposed.　The　results　of　static　tensile　experiments　show　that　the　characteristic　parameter　diagram　of　the　signal　becomes　smoother　after　CEEMD　decomposition.　Finally,　the　method　was　applied　to　the　processing　of　MAE　signal　under　low-cycle　fatigue　experiments.　The　results　show　that　material　yield　and　material　hardening　can　be　found　in　advance　from　the　characteristic　parameter　diagram.　The　experimental　results　show　that　the　proposed　CEEMD　algorithm　has　unique　advantages　in　low-cycle　fatigue　MAE　signal　processing.　©　2021　IEEE.