摘要 |
PURPOSE:To give presence to contents in the case of reading and voice outputting the sentence of a novel or the like by providing a speaker recognition processing part for specifying a conversational sentence within an original and further specifying a speaker in conversation. CONSTITUTION:Characters on the original 1 are optically scanned and made into binarization data at an image data input part 2 and character-recognized and character-encoded at a character recognition processing part 3. A main control part 4 delivers the data of the character recognition processing part 3 to a sentence data storage part 5 to be stored. The conversation in the sentence and the speaker are recognized at the speaker recognition processing part 9 and the tone and the volume of the speaker are set based on the various kinds of the data stored in a data part 11 for deciding the tone and the volume at a speaker data setting processing part 10. The data stored in the sentence data storage part 5 are read by the main control part 4 by each fixed length, delivered to a voice synthesis processing part 6 along with the data of voice setting, converted to voice data and then outputted from a voice output part 7. |