Genetic Tuning of a Laser Pointer Environment Control Device System for Handicapped People Francisco Ch´ avez, Francisco Fern´ andez, Jes´ us Alcal´ a-Fdez, Rafael Alcal´ a and Francisco Herrera Abstract—In this paper we present a new approach for laser- based environment device control systems by laser pointer for handicapped people. The paper proposes the design of a Fuzzy Rule Base System for laser pointer detection. The idea is to improve the success rate of the previous approaches decreasing as much as possible the false offs, i.e., the detection of a false laser spot (since this could lead to dangerous situations). To this end, Genetic Fuzzy Systems have also been employed for improving the laser spot system detection thus reducing the system false offs, that is the main objective in this problem. The system presented in this paper, using a Fuzzy Rule Base System adjusted by a Genetic Algorithm, shows a better success rate, and the most important thing, the not desired false offs are completely avoided. I. I NTRODUCTION Nowadays, people with handicap or chronic illness have less problems to control their home devices than some years ago. Thanks to the research effort in smart homes [1], [2], [3], [4], [5], systems designed for disabled people have been successfully developed. These kinds of systems are adapted to the needs of elderly or disabled people thanks to a computer system which controls the appliances of their homes, by means of which, they can have a normal life. Further, such automatic systems are also able to send information about the behavior of the people in a home in order to avoid dangerous situations. An extensive review on smart homes can be found in [1], showing the great evolution of these kinds of systems in the last years. We can find different kinds of smart home systems in the specialized literature. Park et al. [2] presented a robotic smart house, by means of which disabled people can be assisted by a robot sending orders with body movements. Other researchers have used non-invasive brain-computer interfaces [3], where a brain computer interface is used to control different devices. On the other hand, the work presented in [4] makes use of a robot for helping disabled people to pick up different objects. In this work, the authors also proposed an interesting way to point out the desired objects by means of a laser pointer. The robot should then be able to detect the laser spot on an object in order to pick up this object. Since they represent cheap and easy to handle devices, laser pointers have been used as an indicator element for Francisco Ch´ avez and Francisco Fern´ andez are with the Department of Computer Sciences, University of Extremadura, 06800 Merida, Spain (email: {fchavez, fcofdez}@unex.es). Jes´ us Alcal´ a-Fd´ ez, Rafael Alcal´ a and Francisco Herrera are with the Department of Computer Science and Artificial Intelligence, CITIC-UGR (Research Center on Information and Communications Technology), Uni- versity of Granada, 18071 Granada, Spain, (e-mail: {jalcala, alcala, her- rera}@decsai.ugr.es). controling large displays [6], [7], [8], [9] and now they have been also used to help to disabled people in home device environment control. To this end, an environment device control system for handicapped people has been recently presented in [5]. By means of this system, a handicapped person can use a laser pointer in order to indicate which home device he wants to use. A video camera takes an environment image and sends it to the computer. This image is analyzed with different algorithms for detecting a laser spot. Finally, if the laser spot is on a home device, a KNX/EIB domotic system [10] sends an order for controlling it. Thanks to this kind of systems, people with handicap will be able to control their home devices easily by a laser pointer. The main goal in these types of systems is to detect the laser spot effectively. In [4], the authors deal with this problem by using especial physical filters in the video camera while it is taking the environment photos, with the aim of only capturing the laser spot. However, this is still an open problem that can be addressed by using laser spot detection algorithms based on the original non-filtered images. As we have mention, we also presented in [5] a set of algorithms, that can detect the laser spot effectively on the image obtained by the video camera. Nevertheless, this algorithms have false offs, when a laser spot is detected by the algorithm but the image does not have any laser spot. In this event, a wrong order is sent to the domotic system which could provoke undesirable, dangerous or at least unexpected, situations. In this paper, we present a new approach to detect the laser spot in the environment device control system presented in [5]. It consists of a Fuzzy Rule Based System (FRBS) [14], [15] for trying to improve the success rate in images without laser spot, and to completely avoid the false offs of the previous systems. Since it is a very particular problem, in which only a few training examples are available, the rules comprising this FRBS will be initially obtained by an expert. Even though that this initial system improves the success rate of the previous techniques in [5], it still presents false offs. In any event, we will benefit from the expert knowledge to obtain the initial definition of the FRBS, in order to later improve the FRBS performance by applying a genetic tuning of the Membership Functions (MFs) [11], [12], [13]. The results obtained by the tuned FBRS show a better success rate, and the most important thing, the not desired false offs are completely avoided. This contribution is arranged as follows. In section II, the environment device system for handicapped people by