Python uses Pytesser module to recognize image text

The use of Python pytesser module, originally wanted to do is the image of Chinese recognition, engaged for some time, in the Chinese recognition there are still a lot of problems, here to do record sharing.

　　Pytesser, OCR in Python using the Tesseract engine from Google. is a module of the Google OCR Open source project, which converts the text in the image to text (mainly in English).

　　1.pytesser Installation

Equipment used: Win8 64-bit

　　Pytesser uses the tesseract OCR engine to convert the image to an acceptable format, and then executes tesseract to extract the text information . With Pytesser, you do not need to install the Tesseract OCR engine, but you must first install the PIL module (the graphics library of Python Image Library,python)

Pytesser Download: http://code.google.com/p/pytesser/If not open, can be downloaded through the Baidu network disk: Http://pan.baidu.com/s/1o69LL8Y

PIL Official Download: http://www.pythonware.com/products/pil/

Where PIL can be directly clicked EXE installation, Pytesser without installation, after decompression can be placed in the Python installation folder \lib\site-packages\ under the direct use (need to add pytesser.pth)

　　2.pytesser Source Code

　　By looking at pytesser.py's source code, you can see several key functions:

(1) call_tesseract (Input_filename, Output_filename)

The function calls tesseract external execution program to extract text information from the picture　

  (2) image_to_string (IM, cleanup = cleanup_scratch_flag)

This function handles an image object, so you need to use IM = open (filename) to open the file and return an image object. It calls Util.image_to_scratch (IM, scratch_image_name) to save the in-memory image file as BMP so that the TESSERAC program can handle it properly.

  (3) image_file_to_string (filename, cleanup = cleanup_scratch_flag, graceful_errors=True)

The function directly reads the image file using Tesseract, and if the image is incompatible, it is converted to a compatible format before extracting the text information from the picture.

"""OCR in Python using the Tesseract engine from googlehttp://code.google.com/p/pytesser/by Michael J.T. O ' Kellyv 0.0.1, 3/10/07"""ImportImageImportsubprocessImportutilImportErrorstesseract_exe_name='tesseract' #Name of executable to being called at command lineScratch_image_name ="temp.bmp" #This file must is. bmp or other tesseract-compatible formatScratch_text_name_root ="Temp" #Leave out the. txt extensionCleanup_scratch_flag = False#temporary files cleaned up after OCR operationdefcall_tesseract (Input_filename, output_filename):"""Calls external Tesseract.exe on input file (restrictions on types), outputting output_filename+ ' txt '"""args=[Tesseract_exe_name, Input_filename, Output_filename] proc=subprocess. Popen (args) Retcode=proc.wait ()ifretcode!=0:errors.check_for_errors ()defImage_to_string (IM, cleanup =cleanup_scratch_flag):"""converts IM to file, applies tesseract, and fetches resulting text. If cleanup=true, delete scratch files after operation."""    Try: Util.image_to_scratch (IM, scratch_image_name) call_tesseract (Scratch_image_name, Scratch_text_name_roo T) text=Util.retrieve_text (scratch_text_name_root)finally:        ifCleanup:util.perform_cleanup (Scratch_image_name, Scratch_text_name_root)returntextdefimage_file_to_string (filename, cleanup = Cleanup_scratch_flag, graceful_errors=True):"""applies tesseract to filename; or, if image was incompatible and graceful_errors=true, converts to compatible format  And then applies tesseract.    fetches resulting text. If cleanup=true, delete scratch files after operation."""    Try:        Try: call_tesseract (filename, scratch_text_name_root) text=Util.retrieve_text (scratch_text_name_root)excepterrors. Tesser_general_exception:ifGraceful_errors:im=image.open (filename) text=image_to_string (IM, cleanup)Else:                Raise    finally:        ifCleanup:util.perform_cleanup (Scratch_image_name, Scratch_text_name_root)returntextif __name__=='__main__': Im= Image.open ('phototest.tif') Text=image_to_string (IM)PrinttextTry: Text= Image_file_to_string ('fnord.tif', graceful_errors=False)excepterrors. Tesser_general_exception, Value:Print "fnord.tif is incompatible filetype. Try graceful_errors=true"        PrintValue Text= Image_file_to_string ('fnord.tif', graceful_errors=True)Print "fnord.tif Contents:", text text= Image_file_to_string ('Fonts_test.png', graceful_errors=True)PrintText　　

　　3.pytesser use
　　To load the Pytesser module in the code, the simple test code is as follows:

 fromPytesserImport*im= Image.open ('Fonts_test.png') Text=image_to_string (IM)Print "Using image_to_string ():"PrintTexttext= Image_file_to_string ('Fonts_test.png', graceful_errors=True)Print "Using image_file_to_string ():"PrintText

The recognition results are as follows: Basic can be extracted from English characters, but for some complex points of the picture, for example, I try to identify some English paper images, but the results are not ideal.

　

Since there are many problems in Chinese recognition, we will further study and share them later.

Reference: HK_JH's Column http://blog.csdn.net/hk_jh/article/details/8961449

Python uses Pytesser module to recognize image text