Accurate　identification　of　R-peaks　is　the　basis　of　ECG　correlation　analysis.　In　this　paper,　we　propose　a　method　that　first　enhances　the　features　of　R-peaks　using　difference-square　operations,　and　then　locates　R-peaks　by　threshold　detection,　slope　judgment　and　backtracking　search,　where　the　threshold　is　a　set　of　combined　thresholds　that　are　related　to　R　peak　amplitude,　R-R　interval　and　noise,　respectively.　The　algorithm　was　tested　on　the　MIT-BIH　arrhythmia　database,　and　the　sensitivity　Se　=　99.59%,　while　the　positive　prediction　rate　+P　=　99.71%.　The　method　shows　robustness　with　the　addition　of　different　noise　levels.　The　experimental　results　show　that　the　proposed　algorithm　reduces　the　false　detection　rate　and　increases　the　recognition　rate,　takes　into　account　the　real-time　and　anti-interference　performance,　and　is　suitable　for　the　correlation　analysis　of　ECG.　©　2022,　The　Author(s),　under　exclusive　license　to　Springer　Nature　Singapore　Pte　Ltd.