A detecting device includes a processor and a memory which stores a plurality of instructions, which when executed by the processor, cause the processor to execute: specifying a region of a hand from a captured image; deciding, based on a movement amount of a region specified by the specifying, a first region with a first movement amount and a second region with a second movement amount that is larger than the first movement amount; if a first region and a second region decided by the deciding are captured in a joined state, deriving a region that remains after removing the first region from the captured image; and detecting a predetermined site in an image of a region derived by the deriving.