UAVs　(Unmanned　Aerial　Vehicles)　have　been　widely　used　in　power　line　inspections,　but　low　autonomous　cruise　capacity　of　UAVs　requires　strict　condition　for　operators　and　site　while　landing　during　UAV　power　line　inspections.　This　paper　presents　an　autonomous　landing　control　technique　for　UAVs　when　charging　at　the　electric　towers　based　on　vision　positioning　method.　The　proposed　system　consists　of　three　modules:　visual　processing,　positioning　estimation　and　landing　control.　The　desired　position　and　direction　can　be　obtained　by　ellipse　fitting　and　Shi-Tomasi　corner　detection　method　after　the　landing　pad　was　identified　by　affine　constant　moment　and　Support　Vector　Machine　(SVM)　in　visual　processing　module.　Extended　Kalman　Filter　(EKF)　is　used　for　the　fusion　of　inertial　measurement　with　the　position　data　estimated　by　camera　imaging　principle　in　positioning　estimation　module.　A　control　strategy　is　designed　to　enable　the　UAV　to　land　safely　and　reliably.　The　experimental　results　show　that　the　position　error　of　the　landing　is　within　20cm,　and　the　direction　error　is　within　20　degrees,　which　basically　meets　the　demand　for　autonomous　landing.　©　2017　IEEE.