CIB ocr technical manual (EN)

8. Properties Text-Recognition with deepER

Instead of using Tesseract for OCR it is possible to also choose text-recognition with deepER. The OCR will be calculated on a server. A RESTFUL Service is running on the server, while the client utilizes libcurl in order to send the request.


DataFolder
InputFilename
OutputFilename
Recognize
DeeperURL
DeeperAuthentication
DeeperImageFormat
OcrGrayScaleConversion

InputFilename

This Property is mandatory.  

It is not yet possible to use In-Memory-Processing for the input. 

Property-Name 

Datentyp 

Art 

InputFilename 

String 

Set/Get 

 

This property specifies the name of the input file.  
The following input-formats are supported: 

  • bmp image;
  • tiff image (includes multipage tiff);<
  • jpeg image;
  • png image;

OutputFilename

Property-Name 

Data-Type 

Type 

OutputFilename 

String 

Set 

 

This property specifies the name of the out putfile.  
The Outputformat is fixed to hOCR. 

 

Syntax 

OutputFilename=<name> 
<name>: name.ext  

 

Example 

OutputFilename=Rechnung.html

Recognize

Property-Name 

Data-Type 

Type 

Recognize 

String 

Set 

 

Syntax  

Recognize=<Value>
<Value>: OcrRecognizerWithDeeper 

default=OcrRecognizer 

 

Example 

Recognize= OcrRecognizerWithDeeper 

DeeperURL

Property-Name 

Data-Type 

Type 

DeeperUrl 

String 

Set 

 

Syntax 

DEEPERURL=<Value>
<Value>: http://localhost:5000 

default= http://localhost:5000 

 

Example 

DeeperUrl = http://graphix:5000 

DeeperAuthentication

Property-Name 

Data-Type 

Type 

DeeperAuthentication 

String 

Set 

 

Syntax 

DEEPERURL=<Value>
<Value>: User:password 

default= “” 

 

Example 

DeeperUrl = Franz:TopSecret 

DeeperImageFormat

Property-Name 

Data-Type 

Type 

DeeperImageFormat 

String 

Set 

 

Default value (if not set) is PNG. Possible values: JPG (JPEG) / PNG / Smallest.  

CIB ocr converts input image into the requested format, before sending it to the deeper server. 

If DeeperImageFormat is set as Smallest then CIB  ocr converts the input image into both: PNG and JPG and the smallest representation will be sent to the deeper server for recognition . 


OcrGrayScaleConversion

Property-Name 

Data-Type 

Type 

OcrGrayScaleConversion 

String 

Set 

 

Syntax 

OcrGrayScaleConversion=<Value>
<Value>: 0|1 

Default =1 

 

Example 

OcrGrayScaleConversion = 0