For　the　enormous　number　and　the　limited　energy　of　network　nodes　in　the　wireless　video　sensor　networks　(WVSN)　environment,　to　fulfil　the　complicated　tasks,　multiple　sensor　nodes　should　collaborate　with　each　other.　A　cooperative　multi-object　tracking　method　for　Wireless　Video　Sensor　Networks　is　proposed　in　this　paper.　The　proposed　method　is　focused　on　the　solution　of　cooperative　multi-object　tracking　among　multiple　sensor　nodes　when　an　object　leaves　the　view　field　of　the　tracking　node.　The　main　contributions　of　our　proposed　method　are　that:　(1)　the　sensing　model　of　a　video　sensor　and　Kalman　filter　is　utilized　to　achieve　optimal　sensor　selection.　(2)　Projective　Invariants　are　employed　to　integrate　information　from　the　related　nodes.　The　experimental　results　show　that　the　proposed　method　is　effective　for　resolving　the　problem　of　tracking　relay.