Design of Image Retrieval System Based on Speech Recognition

Abstract

Nowadays cross-media data mining and retrieval becomes an urgent problem in order to use multimedia resources more efficiently. This paper presents an overall design method on an image retrieval system based on embedded development. On the basis of speech-recognition technology, the voice command is converted into the text which is later to be as the keywords when searching image. In this way, our system allows users to just say can reach the purpose of image retrieval. The tests prove we successfully achieve the image retrieval after inputing voice-command. The integration of speech and image recognition technology can greatly free people's hands and provide a great convenience to users, especially for the illiterate or some disabilities.