Hand-hold character recognizing instrument based on SMS

Abstract

This invention relates to one handle character identification meter based on SMS composed of hardware part and software part, wherein the hardware part is imbedded with micro processor, power module, image collection module, SDRAM, FLASH, series communication module and short message communication module; the power module is to provide power for identification device; image collection module is to collect image data; SDRAM runs operation system to read image data; series communication module connects the imbed micro processor and short message module to send the final result to cell phone; software part is composed of guide load program, windows CE operation system core and application program.

Description

(1) Technical field: [0001] The present invention is a hand-held character recognition device based on SMS. SMS is the abbreviation of Short Message Service. The recognition device involves image data collection, digital image recognition and wireless digital communication technology, and belongs to pattern recognition and detection devices. . (two) background technology: [0002] With the development of information society, image processing and machine vision technology are more and more widely used in various industries. Character recognition technology has developed rapidly in recent years, and has broad application prospects in logistics information, instrument readings, license plate recognition, bill recognition, ID card identification, etc., in order to achieve the purpose of improving efficiency and reducing labor costs. [0003] For the recognition of characters in images, most of the current research objects are for single characters. Therefore, for the collected ...

AI Technical Summary

Problems solved by technology

These methods lack versatility. When the brightness of the environment changes greatly, the effect of binarization also changes greatly, which directly affects the subsequent character recognition.

There is also the use of the region growing algorithm and connectivity characteristics to directly binarize the characters and extract them from the background. This method does not consider prior knowledge and has certain versatility, but the calculation process is quite time-consuming and is not suitable for embedded applications. in the system

[0004] At present, character recognition methods mainly include methods based on template matching, methods based on character structure and methods based on neural networks. These existing image recognition methods have more or less certain limitations. A method that works very well, changing to another environment recognition effect may be very unsatisfactory

Some methods with certain versatility and good recognition effect often have a large amount of calculation and are difficult to apply in real time

Images