This is a repository for docker image of processing files with ABBYY Cloud OCR, using Python 3.7.
It is a Python code that was gotten from Quick start with OCR SDK for Python.
This is not an asynchronous code, so it hangs a little bit waiting a return from the ABBY Cloud.
IMPORTANT: The are a bunch of API methods available but this code takes care only of processImage method and processTextField.
First of all you need an account at ABBY Cloud OCR SDK and also an application configured for your submission.
After creating your account, create an application and save its name. A password will be sent to you through mail.
After cloning this repository, being on the main folder, run:
docker build -t abbyy-ocr-process .On your environment, code below is for Linux, export three important variables as follows:
- ABBYY_SERVER_URL - URL of the ABBYY OCR Server where you have registered you account.
- ABBYY_APPID - The application ID that you registered.
- ABBYY_PWD - The application password that was sent to you by ABBYY.
export ABBYY_SERVER_URL="https://cloud-westus.ocrsdk.com"
export ABBYY_APPID="<YOUR_APPLICATION_ID>"
export ABBYY_PWD="<YOUR_APPLICATION_PASSWORD>"There are optional arguments that can be sent with shortcut letters, as follows:
| Optional Argument | Description |
|---|---|
| -l or --language |
Specifies the recognition language. Default: English Supported languages |
The following arguments represent a mutual exclusive group and just one of them can be used, and in fact, must be used for choosing which API method should be used: processImage or processTextField
| Optional Argument | Description |
|---|---|
| -image This is DEFAULT. |
Specifies the operation as image which calls the processImage API method. |
| -textField | Specifies the operation as textField which calls the processTextField API method. |
The following arguments represent a mutual exclusive group and just one of them can be used, and in fact, must be used for processing the OCR request through processImage API method.
| Optional Argument | Description |
|---|---|
| -txt This is DEFAULT. |
The recognized text is exported to the file line by line from left to right. E.g. if the text was originally put in columns, the first lines of every column will be saved, then the second lines, etc. Please take into account the fact that in this format only text will be saved. No images or barcodes will remain in the output file. If you want to save the barcode recognition results in the exported file, use the txtUnstructured format. |
| -txtUnstructured | The exported file contains the text that was saved according to the order of the original blocks. |
| -rtf | Exports the result into RTF format. |
| -docx | Exports the result into DOCX format. |
| -xlsx | Exports the result into XLSX format. |
| -pptx | Exports the result into PPTX format. |
| -pdfSearchable | The entire image is saved as a picture, the recognized text is put under it. |
| -pdfa | The file is saved in the PDF/A-1b format, with the entire image saved as a picture, and recognized text put under it. |
| -pdfTextAndImages | The recognized text is saved as text, and the pictures are saved as pictures. |
| -xml | Exports the result into XML format. |
| -xmlForCorrectedImage | The same as xml, but all coordinates written into the output XML file relate to the corrected image, not the original. |
The following arguments represent a mutual exclusive group and just one of them can be used, and in fact, must be used for processing the OCR request through processTextField API method.
| Optional Argument | Description |
|---|---|
| -allTextTypes This is DEFAULT. |
Specifies text type as normal,typewriter,matrix,index,ocrA,ocrB,e13b,cmc7,gothic to be used by processTextField API method. |
| -normal | Specifies text type as normal to be used by processTextField API method. |
| -typewriter | Specifies text type as typewriter to be used by processTextField API method. |
| -matrix | Specifies text type as matrix to be used by processTextField API method. |
| -index | Specifies text type as index to be used by processTextField API method. |
| -ocrA | Specifies text type as ocrA to be used by processTextField API method. |
| -ocrB | Specifies text type as ocrB to be used by processTextField API method. |
| -e13b | Specifies text type as e13b to be used by processTextField API method. |
| -cmc7 | Specifies text type as cmc7 to be used by processTextField API method. |
| -gothic | Specifies text type as gothic to be used by processTextField API method. |
You must have set your environment variables before starting the container.
The following example, code below is for Linux, submits all files inside input folder, of host that is mapped to container, and request its return to be a XLSX using Brazilian Portuguese as the recognition language.
Run the following:
docker run -e ABBYY_SERVER_URL="$ABBYY_SERVER_URL" \
-e ABBYY_APPID="$ABBYY_APPID" \
-e ABBYY_PWD="$ABBYY_PWD" \
-v $(pwd)/input:/usr/app/input \
-v $(pwd)/output:/usr/app/output \
ricardosouzamorais/abbyy-ocr-process \
-l "PortugueseBrazilian" -xlsxThe following example, code below is for Linux, submits all files inside input folder, of host that is mapped to container, and request its return to be an XML (only available output format for this API method) using Brazilian Portuguese as the recognition language.
Run the following:
docker run -e ABBYY_SERVER_URL="$ABBYY_SERVER_URL" \
-e ABBYY_APPID="$ABBYY_APPID" \
-e ABBYY_PWD="$ABBYY_PWD" \
-v $(pwd)/input:/usr/app/input \
-v $(pwd)/output:/usr/app/output \
ricardosouzamorais/abbyy-ocr-process \
-l "PortugueseBrazilian" -textField