Sign language is a method of communication, especially with people in special needs such as those that are deaf and dumb, but this language is not easily understood by everyone. Many articles and researches have focused on learning hand language and converting hand signals into meaningful signs, by using signal processing methods to translate or to convert sign language to speech or texts, for example, showing the number or the text on the screen or converting it to a spoken language. In this paper, an image processing technique has been proposed, to extract the hand signal feature of English numbers and convert them into written text. Specific gloves were used to simplify the extraction of features using two symbols- a circle and a triangle that was printed on each glove. An algorithm was applied to each image with its feature extracted by MATLAB programs. The process was applied to each PNG image by converting it to a binary image and using object detection, sobole filter for edge detection, image resizing, calculating the number of circles and triangle. The correlation between each image feature was measured to identify the specific sign language and then to convert it to texts or audio.
CITATION STYLE
Chabchoub, A., Hamouda, A., Al-Ahmadi, S., Barkouti, W., & Cherif, A. (2020). Hand Sign Language Feature Extraction Using Image Processing. In Advances in Intelligent Systems and Computing (Vol. 1070, pp. 122–131). Springer. https://doi.org/10.1007/978-3-030-32523-7_9
Mendeley helps you to discover research relevant for your work.