Semantic Annotation of Images with Text and Sound for Visually Impaired

Sarika Jain; Siddharth Prasad; Akhilesh Kumar Lodhi

doi:10.37591/joosd.v5i1.1483

Abstract

Accessing appropriate information is the biggest challenge for human and specifically for visually impaired people, and when the information to be accessed is multimedia, it poses a big problem. There is a need of annotating (labelling) the images with text and sound to make it easier for visually impaired people to understand what the image is about. In addition to the ease of understanding, semantic annotation of images can make the semantic search of images simpler, provides ease in visualization, provides a platform for education toolkit, increases tourism by making historical pictures self-speaking, and many more such applications. Semantic Annotation binds ontologies with documents via metadata. A variety of tools have been developed to help visual impaired individuals to interact with the world around them but because of accessibility and cost, they are unapproachable. HELPI has been developed as a powerful tool for annotating the pictures using ontology-based multimedia concepts. Text and sound will be added to pictures using multimedia ontology(m-owl), the meta data thus created is converted in mpeg7 format. “HELPI” makes pictures to speak and help visually impaired people in order to understand and enjoy the content of pictures in a better way.