A region in a captured image is set as a signal detection region, and signals from an CMOS sensor of the rolling shutter exposure type are detected only in an exposure period corresponding to the signal detection region to produce an evaluated value, after which a focusing lens is moved to a focusing lens position for acquiring a next evaluated value. Since the signal detection region is a partial region in the captured image, a long detection quiescent time is kept between signal detection periods corresponding to every two horizontal synchronizing periods. By moving the focusing lens in the detection quiescent time, the evaluated value is acquired and the focusing lens is moved in each horizontal synchronizing period.