Method, apparatus, and non-transitory medium for generating a synthetic image from a series of captured images

An image processing apparatus includes: a specific sound detecting section detecting a specific sound generated during motion in a specific motion of a target object included in an image group including a plurality of images which continues in a time-series manner; a valid sound range setting section setting, as a valid sound range, a range in the time axis for determination of whether the specific sound is valid on the basis of a user manipulation; and a selecting section selecting a predetermined number of motion transition images representing transitions of the specific motion from a selection range of the image group in the time axis, with reference to a detection position of the specific sound in the time axis, in a case where the specific sound is detected in the set valid sound range.