In　current　power　inspection　work,　a　large　number　of　defects　need　to　detect　by　observing　visible　images.　In　order　to　cope　with　this　problem,　a　lot　of　researches　are　carried　out　to　realize　automation　of　vision-based　inspection.　In　this　paper,　with　the　aim　of　providing　a　good　starting　point　for　researchers　interested　in　this　field,　an　extensive　literature　review　is　conducted.　Firstly,　the　source　of　visible　image　for　power　inspection　is　presented,　the　items　and　defects　of　visible　inspection　are　summarized,　and　visual　inspection　of　visible　image　is　divided　into　two　types　of　issues:　image　object　detection　and　image　distance　measurement.　Then,　the　two　types　of　issues　are　reviewed,　the　development　trend　of　intelligent　visual　inspection　of　visible　image　is　analyzed　with　experimental　data,　and　the　application　prospect　of　deep　learning　is　discussed.　On　this　basis,　the　challenges　and　possible　solutions　for　further　application　of　deep　learning　are　expounded.　Finally,　the　paper　is　concluded　with　an　outlook　for　the　future　of　this　field　and　an　proposal　of　potential　next　steps　for　implementing　the　concept.　©　2020,　Power　System　Technology　Press.　All　right　reserved.