Monday, March 29, 2010

Digital Cameras Input language skills

Digital Cameras Input language skills

with a digital camera for fast text entry, it is very simple and is made into a photo spread into the computer, and then recognition software to convert plain text file format can be.


Whenever your boss throw a thick stack of files, so you are typing a word for word the computer, do you feel is not going to collapse? Have you ever thought, a digital camera you can make it better and more economical to "typist"?


Today, digital cameras have large numbers into ordinary homes. For ordinary users, the use of digital cameras to shoot photos, it is only right and proper thing. However, with digital cameras to input text, I am afraid that most people in the eyes of a little fantasy. Now, let's first look at the Du Men capacity unique skills - with digital cameras to input the text of the real "Kung Fu Hustle."


shooting


with a 300-million or more pixels, more than 3 times optical zoom digital camera that filmed the original manuscript. When shooting should note: the manuscript should be flat, the camera should end flat, light to gentle uniform. Can scatter light in the corridor or window, the three primary colors can also be taken under fluorescent lamp eye protection.


taken correctly identify clearly the key to the later, 300 million or more pixels can guarantee for identification after shooting the text is large enough; 3 times optical zoom when shooting so that the manuscript in each of lines covered the screen. I have been comparing the 300-megapixel shot than shooting with 2 million pixels recognition rate is significantly higher.


Do not use flash, light is too strong, shot down but the writing is not clear.



conversion


in the computer's hard drive to create a folder, the photographed images into one. However, these images contrast is very small, the entire screen is gray; and a variety of digital camera image format is different, OCR software (OCR software) may not be able to support, so we are advised to use ACDSee or Windows built-in "Paint & rdquo ; program conversion.


writer is a Windows comes with the "Paint" program to convert the. First, the order click on "Paint → File → Open" in the "Look In" to find the pictures, open and click "Save As" in the "Save as type" Select " monochrome bitmap. "


Click "Save\Converted obtained new pictures follow the original name, changing the extension to ". Bmp\


identification


run OCR software (which can be downloaded from the Internet, the author uses Tsinghua TH-OCR2000 Millennium Professional Edition), then click "File → Open" in the "Look" and find your pictures. Click on "order" drop-down menu appears, if the need for full recognition, you can click on "all recognition\The upper part of the software window after shooting, the converted original image, the lower part of the text is identified. Recognition followed automatically created after the original name of the txt file, saved in the original file folder.


If only identify part of the text, press and hold the left mouse button in the need to identify parts of the text pulled out the box, there a yellow block, click on "identification" can be.


finishing


created at the beginning of the file folder to find the extension txt plain text file file, this is the information needed for identifying the wrong word to modify, especially in the text emphasis added and underlined text recognition rate is still relatively low. At the same time in the file folder, there are many extensions chr, trc, sim files, which are generated when OCR recognition temporary files can be deleted.


application skills view includes more and "digital cameras, text entry techniques" related to the article, for the first capacity in this area is still quite a lot of articles, if you think this article is the wrong place, or Do you have a unique insight, you are welcome to express their views to the computer knowledge and recommendations of the Forum. If you like to learn computer knowledge, please keep in mind the only site domain name

No comments:

Post a Comment