Development of a ZigBee based virtual eye for visually impaired persons

Abstract

This research aims to develop a visual aid to the visually impaired person with the following functionalities: i) Localization of the person within the local environment. ii) Location of obstacles within the indoor system. iii) Generation of voice commands to guide the person for his movements without hitting the obstacles. The indoor system is proposed to be designed using ZigBee, sonar and specially designed software systems for voice processing and command generation and a microcontroller for central processing along with other hardware systems required. For indoor localization ZigBee protocol is used with the optimized Received Signal Strength Indicator (RSSI) values. An optimization algorithm is developed for localization using RSSI values for the customized environment. In this module, the device is kept with the visually impaired person. The device uses ZigBee to know its location. In an indoor environment, like house, from one room to other room or in a semi indoor environment from one floor to other floor if the blind person wants to make his movements, the environment must be equipped with a minimum of two fixed listener ZigBee modules with known location coordinates. When the blind person enters in to any of the indoor locality, the ZigBee in the device starts communicating with other two ZigBee protocols which are already fixed in the selected locations in the indoor environment. The three ZigBee modules exchange information through RSSI values. From the RSSI values the distance between the ZigBee modules have been estimated using the look up table of data. The estimated distance information provides a way to find the indoor location of the blind person. The signal attenuation parameter η is sensitive to the environmental variation and the initial signal strength A is a function of time. Thus, attenuation parameters η have to be calculated for various distances and at various conditions. After estimating the value of no loss signal A and η, the look up table is referred for the nearest possible distance using maximum likelihood estimation. This helps in fitting the RSSI value to the exact distance in the look up table.