Voice and single-pen mixed computer Chinese character input method

Abstract

The invention discloses a composite voice and single pen computer Chinese character input method. The method comprises the steps as follows: firstly, a voice signal of the Chinese character to be input is transmitted to a voice identifying module to be identified as Chinese pinyin; at the same time, an initial stroke of the Chinese character to be input is transmitted to a handwriting identifyingmodule by a tablet for identifying a handwriting shape as a stroke number; the pinyin and the stroke number are combined as an intermediate text, and an intermediate text processing interface is displayed; a whole word, a whole sentence or a whole paragraph is input and then is checked and ratified; and finally, a comparing and optimizing module contrasts the input with a grammar, corpus and knowledge base again and gives best input content. The invention reduces coincident codes greatly, has faster input speed than the prior art and has high identification accuracy and low operand.

Description

technical field [0001] The invention relates to the technical field of computer voice recognition and Chinese character input, and relates to a computer Chinese character input method combining voice input and single-stroke handwriting. Background technique [0002] Current computer Chinese character input method is mainly divided into voice, keyboard and three kinds of input modes of handwriting, and these methods have following defect: [0003] 1. There are too many repeated codes in the current voice input method, which requires a large number of knowledge bases to be compared and calculated, which requires a large amount of calculation and low accuracy. [0004] 2. The current voice input method mostly adopts continuous voice input, which often makes mistakes in distinguishing signals and noises and segmenting syllables. In addition, real-time multi-level comparison operations are performed, resulting in relatively insufficient computing power of the computer. [0005] ...

AI Technical Summary

Problems solved by technology

[0003] 1. The current voice input method has too many repeated codes, resulting in the need to compare and calculate a large number of knowledge bases, with a large amount of calculation and low accuracy

[0004] 2. The current voice input method mostly uses continuous voice input, which often makes mistakes in distinguishing signals and noises and segmenting syllables, coupled with real-time multi-level comparison calculations, resulting in relatively insufficient computer computing power

[0005] 3. The current handwriting input needs to input the whole Chinese character, and the input is slow

[0006] 4. Chinese character keyboard input is not humanized enough

[0007] At present, there have also been computer Chinese character input methods that integrate voice and handwriting, such as the public number CN1549244 titled "Recognition Method and System for Integrating Voice and Handwriting Input" and CN101082836 titled "A Chinese Character Integrated Voice Input and Handwriting Input Function" The invention of "input system", but the input of Chinese characters is limited to the input of the entire Chinese character, the input speed is not good enough, and the calculation of Chinese character recognition is heavy.

Images