Lip contour detection based AAC device using morse code

Various kinds of technologies are being developed to improve the communication opportunity of people with Verbal challenges and lesser motor skills. Augmentative and alternative communication (AAC) devices are helpful for non speaking individuals; they use other form of communication such as non-descriptive noises or physical gestures. These noises or gestures will be converted by AAC devices and communicated to the outside world. Lip contour detection has always been an area of interest and several HCI applications are developed based on this. Lip detection is a task difficult to accomplish because of the variation in lip shapes and sizes for different individuals. In this paper, an AAC device that makes use of lip contour detection is proposed. An algorithm for recognition of the lip status is used, which includes the segmentation of the background of the person of interest, this can resolve a major issue of the system detecting other unwanted images or faces on the background of the person of interest. This is followed by face detection of the individual, localization of mouth using region of interest and a hybrid method for lip contour detection is employed. After the detection of the lip contour, the lip status is obtained i.e. the shape of the lips are detected. Depending on the lip shape corresponding Morse code is allotted and this is given to the hardware for processing and displaying the word or alphabet. In conclusion, an AAC device is designed without the need of wearing any appendages by the disabled individual.