TEXT UNDERSTANDING BY PICTURE-IN-PICTURE. AND ROBOT THAT ATTACH SPEECH RECOGNITION.

摘要

The subject innovation relates to a document image recognition and speech recognition, robotics, specifically to a system that offers a variety of document information, the audio information whether or not, storage, and output the information more conveniently. In the information processing related to image processing, document processing, audio processing, and the output storage section, consisting of a main processor document image recognition and speech recognition robot.; According to the present invention in a robot provided to recognize the image document information and audio information: receiving the input document image by the image store to read the image processing, the document image creating a clear image by removing noise change as a black and white image, and the document the document processor for storage by converting the image into text, through a microphone recognizing the external sound and filter to create a clean speech comparing the voice and audio processing unit for driving out the natural language voice, outputting the converted text documents and voice, and outputs the file is the output, a storage unit for storing a DB (database), images, documents, sound processing and the output, a document recognition unit and a robot equipped with speech recognition unit is configured by including a main controller for controlling the stored image is provided.