Video　processing　software　is　often　used　to　remove　specific　moving　foreground　from　a　video.　Existing　forgery　algorithms　for　detecting　this　type　of　tampering　generally　suffer　from　inefficiency　and　are　not　effective　for　the　forged　videos　under　complex　background.　To　address　these　problems,　we　propose　a　novel　forgery　detection　algorithm　for　detecting　video　foreground　removal.　The　algorithm　first　calculates　the　energy　factor　(EF)　of　each　frame　to　identify　forged　frames.　An　adaptive　parameter-based　visual　background　extractor　(AVIBE)　algorithm　is　then　designed　to　detect　suspected　regions　from　the　forged　frames　determined　in　the　first　stage.　After　eliminating　false　detection　by　calculating　the　difference　of　EF　between　suspected　regions　in　the　forged　frames　and　the　corresponding　regions　in　the　authentic　frames,　the　algorithm　finally　locates　the　tampering　traces.　The　experimental　results　show　that　our　proposed　algorithm　has　higher　computational　efficiency　and　accuracy　as　well　as　better　robustness　than　those　of　previous　algorithms.