Ubiquitous computing environments require a user interface that is not only intelligent but also natural and intuitive. Researchers are working on various user interfaces based on gesture recognition and/or speech recognition, but such interfaces still have practical problems such as large computation time and recognition errors. FreeMouse introduced here is both a spatial mouse for controlling physical objects and a TV pointer for manipulating virtual objects. For instance, one can use FreeMouse to perform now familiar actions such as point-and-menu, point-and-command and pick-and-drop on/between appliances in the physical space and objects on the screen. We also introduce here an online signature verification system for FreeMouse, that allows users to sign on the screen. Although the signature verification system may better be described in a separate paper, it is decided to be included here because it is one of the most demonstrative usage example of FreeMouse as a TV pointer that differentiates it well from other types of input devices for TV. The first prototype of FreeMouse was evaluated in conjunction with a smart-home project, and was proved to be easy to learn and intuitive to use.