Click-and-read and question answering method and system based on image recognition and tracking technology

Abstract

The invention discloses a click-and-read and question answering method and system based on image recognition and tracking technology. The method comprises a click-and-read mode and a question answering mode. When the click-and-read mode is set, if a user clicks a triggering area on a page by a click-and-read pointing component, a pre-recorded click-and-read audio corresponding to the trigger areais played. When the question answering mode is set, a pre-recorded question audio corresponding to a current page is played, whether the user clicks the triggering area by the click-and-read pointingcomponent is judged, and a judgment result voice is returned. The system comprises a camera device, a click-and-read question answering device, a storage device and an audio playback device. Accordingto the invention, a book reader can be helped to obtain a better reading experience, and the degree of knowledge grasp of the book reader can be checked through questions and answers.

Description

technical field [0001] The present invention relates to a point reading method and system, in particular to a point reading question answering method and system based on image recognition and tracking technology. Background technique [0002] The reading pen is a new generation of intelligent reading and learning tools developed by using optical code recognition technology and digital voice technology. It can simultaneously realize many functions such as point reading, repeat reading, follow-up reading, recording, and entertainment. The technical principle is to first encode the content of the book through OID, and use special printing technology to print the code on the special book, and then the user can scan the book code with a point reading pen to recognize and play the corresponding voice. When the user uses the reading pen to scan the content of the book, the sound content emitted by the reading pen is combined with the content on the scanned book, and augmented reali...

AI Technical Summary

Problems solved by technology

[0004] The system and method combine OCR technology, so that the point-and-click system can be applied to point-reading of ordinary books, and the applicable area is greatly improved, but it has the following problems: 1) It is necessary to recognize the depth of the book page and the camera, and the depth of the finger from the page , the depth recognition error of the monocular camera is very large, and the recognition accuracy is not ideal; the depth recognition accuracy of the binocular camera will be higher, but the recognition cost is higher; 2) The voice of the point-to-read playback is extracted from the page through OCR technology After the text in the image, the voice synthesized by the machine, the OCR technology cannot recognize the image, and the highest text recognition accuracy rate can only reach about 95%; the voice synthesized by the machine is unnatural, and the affinity is poor; 3) Due to the occlusion of the book by the hand, the image collected in real time is incomplete, and the recognition conversion from image to text will have errors

Images