Hand input system and method

Abstract

The invention relates to the technical field of electronic equipment input, in particular to a hand input system and method. The hand input system comprises a terminal device and a cloud computing server, wherein the terminal device is provided with an input unit, a local identification unit, an identification result fusion unit and a display selection unit; and the cloud computing server is provided with a cloud identification unit. On the one hand, the invention reduces the requirement of the hand input method on the terminal hardware to ensure that the hand input can be used on a lower-grade terminal with a network device, and enlarges the use range of the hand input, on the other hand, the invention improves the identification rate and the identification speed of the hand input through strong computing capability and storage capability of the cloud computation.

Description

technical field [0001] The invention relates to the technical field of electronic device input, in particular to a handwriting input system and method. Background technique [0002] The handwriting recognition input method is currently used very commonly, whether it is on a desktop or an embedded handheld device such as a mobile phone, GPS navigation device, e-book, etc. Recognition accuracy, recognition efficiency and recognition dictionary size are three important technical indicators. Improving recognition accuracy and recognition efficiency and reducing the size of recognition dictionary are the goals of handwriting input method technology developers. However, due to the different devices that run the handwriting input method, there are great differences. Some relatively low-end hardware devices cannot meet the basic requirements of handwriting recognition, such as RAM size, CPU frequency, etc., which limits the use of handwriting input methods on these devices. [0003...

AI Technical Summary

Problems solved by technology

However, due to the different devices that run the handwriting input method, there are great differences. Some relatively low-end hardware devices cannot meet the basic needs of handwriting recognition, such as RAM size, CPU frequency, etc., which limits the use of handwriting input methods on these devices.

[0003] In the existing technology, in order to enable the handwriting input method to be used on relatively low-end devices, the algorithm is constantly optimized. At present, the main technical means are: the use of reduced recognition character sets, such as only supporting 6763 characters of GB2312-80, and cannot provide Recognize all characters in the GB18030 character set; in addition, use a relatively simple and low-precision recognition method to reduce the amount of calculations for recognition and the volume of the recognition dictionary, such as using a single template, a simple Euclidean distance classifier, etc., but doing so The result makes the recognition rate drop a lot, greatly reduces the efficiency of handwriting input, and also affects the user experience

Images