It is very difficult to mechanize tomato harvesting because identifying a target tomato which is partly covered by leaves and stalks is not easy. This research was conducted to develop tomato harvesting robot which can identifying a target tomato, determining its dimensional position, and harvesting it in a limited time. Followings were major findings in this study. The first visual system of the robot was composed of two CCD cameras, however, which could not detect tomato not placed on the center of lens and partly covered by leaves or stalks. Secondary visual device, combined with two cameras and pan tilting was designed which could decreased the positioning errors within $pm$10mm but still not enough for covered tomato by any obstacles. Finally, laser detector was added to the visual system that could reduce the position detecting errors within 10mm in X-Y direction and 5mm in Z direction for the covered tomatoes.