Interpreter sign language for communication hearing impaired using image processing

Abstract

It is presented a computational tool based on image processing conceived to assist people with hearing disability. The tool is able to capture a sequence of images corresponding to the gestures the user makes with his/her hands to perform any of the signs used for nonverbal communication. The tool submits this information to a series of processing steps in order to properly identify which gesture is and to present an equivalent, textual information. The results show that the signs for the American Sign Language, the worldwide most employed language and base for the Colombian language were identifi ed correctly and in a variety of situations. Moreover, the tool allows the user to create their own signs, thus being not only an assistant tool, but also a support environment for further research in the area.