Provided is a method for more accurately correcting position coordinates of a point on an object to be imaged, the coordinates being identified based on values detected by linear scales. A visual field is moved to a measurement point defined on a recessed portion formed on a calibration plate, and an image is captured (step S13-1), edges are detected from an image of sides of the recessed portion (step 313-2), an intersection of the edges is calculated (step S13-3), values of the intersection as actually measured by the linear scales are saved (step S13-4), and position coordinates of the point on the object to be imaged as detected by the linear scales are corrected by using a true value and a difference.