PearOCR Free OCR image to text, can also be used offline

Everyone knows that if text appears on a picture, it cannot be directly To copy or edit text content, you must scan through the OCR optical character recognition technology, and then convert the result into a plain text format for output. The extension function of Project Naptha was introduced before. The text can be copied and selected directly. Later, there are many free OCR tools that support Chinese, such as EasyScreenOCR and NewOCRFree Online OCRetc., and now The camera lens of the mobile phone also has a built-in OCR effect, which is very convenient to recognize the text content after shooting directly.

“PearOCR” recommended in this article is a very powerful free OCR text recognition tool this service is completely free, and there is no limit on the number of times it can be used. It supports traditional Chinese, Simplified Chinese and English content, the biggest feature of is that PearOCR can complete all calculations offline no need to use any third-party service APIs, and no need to upload pictures, which also ensures that user data will not In the process of leakage or privacy and security concerns.

PearOCR uses a self-developed lightweight recognition engine, and the model obtained through deep learning training has very powerful text recognition capabilities. The entire recognition process is completed on the web page. Just select the local picture or directly read the scrapbook ( In the case of copying pictures), text recognition can be performed, and the Chinese and English content of the pictures can be converted into plain text or PDF format.

As mentioned earlier, PearOCR can be used offline, and it supports the PWA standard (Progressive Web App). As long as the website has been opened before, it can be used normally even if there is no connection. Text recognition function, or install it to the desktop by application.

PearOCR text recognition
https://pearocr.com/

Use teaching

STEP 1

to open the PearOCR text recognition tool, select from the “ recognition language ” below The language to be used can be switched to Traditional Chinese, Simplified Chinese or English.

STEP 2

Then directly drag the picture to be recognized to the PearOCR website, or select “local picture” or “ read from below Clipboard ” two functions, the former is to choose to upload the recognized picture file, if the picture has been copied to the clipboard, you can use the latter to directly let the web page read the contents of the clipboard (it will pop up a prompt message that requires partial permissions for the website) .

STEP 3

The browser will start scanning and recognizing the text of pictures uploaded by users. It will not take too long and the speed will be quite fast.

After OCR recognition is completed, the detected text will be displayed on the right side, and the plain text on the right side can be copied and used directly.

STEP 4

Click on the option “Display in plain text ” at the bottom right, and the scanned text will be displayed in plain text format, of course it is possible Because of the size, font or color of the text, the recognition error or the mistaken imprinting of other text is caused, and it is necessary to manually calibrate it once after use to avoid the occurrence of typos.

In addition, the option below also has a “Display in a table format” which is also very interesting. It will display the recognized characters on the picture The text and the confidence level are also convenient to take out a certain paragraph of text, and this value can be adjusted through the “ set ” option in the lower right corner.

STEP 5

Click the PearOCR setting function in the lower right corner, you can set “ Confidence Threshold ” or text background color, text color and other parameters , Save it after setting, and click “ Re-identify all pictures ” to help get more accurate results. If the scanned results are very different, you can try to adjust these parameters and try again.

STEP 6

PearOCR has a very useful function in the lower right corner of “ Display the picture under the text “, which can overlap the original under the recognition result The picture allows the user to know where it is immediately when accessing the text content.

In addition to directly copying the plain text content recognized by OCR, you can also click “ Export all images to PDF” to convert the recognized text It is in PDF format. By default, the recognized text and images will be exported together, and the original image can be attached to the next page, which is convenient for viewers to view and compare, and copy the plain text directly.