VISUAL RECOGNITION POSITION CORRECTING METHOD

PURPOSE:To always capture a sensing object in a visual field, and enable continuous mounting, by sequentially correcting an instruction position at the next instruction point, on the basis of results of position correction detected by visual recognition at each instruction point. CONSTITUTION:A mounter obtains position differences x, y between a land 2b on which a parts is actually mounted and an instructed land 2a while the parts is mounted on an instructed position, and sets the differences as the visual recognition position correcting amount. When the land 3 on which a parts is to be mounted is in the outside of the visual field of a visual recognition area 6a at the time of instruction, the visual recognition position 6a is moved by using the previously obtained visual recognition correcting amount x, y, and a new visual recognition position 6B is set. A parts is mounted on the position of the land 3b obtained by the position 6b, and a new visual recognition correcting amount is obtained. Then the above process is repeated. Thereby the midway interruption caused by visual recognition impossibleness is avoided, and continuous mounting is enabled.