PrizmDoc Viewer v13.25 Release - Updated
PrizmDoc OCR API / How to OCR an Image
In This Topic
    How to OCR an Image
    In This Topic

    Introduction

    This section describes how to use the PrizmDoc OCR REST API to perform an OCR operation on an image running the OCR Reader Worker in a Docker container.

    See the OCR Reader Sample on our GitHub page for details on how to use the PrizmDoc OCR REST API.

    Step 1: Run OCR Reader Worker

    Run the OCR Reader Worker as a Docker container as described in the How to Run a Worker topic.

    Step 2: Upload Your Source Document

    • Upload the source document that you want to convert.
    • This can be a document of any format supported by the PrizmDoc OCR REST API.
    • In response to this request you will receive a file ID that is used to reference the source document in later requests.
    • See the Work File API for more details about uploading work files.

    Example

    POST http://localhost:18681/PCCIS/V1/WorkFile?FileExtension=png
    Content-Disposition: form-data; name="image"; filename="testimage.png"
    Content-Type: image/png
    [binary data]
    
    200 OK
    Content-Type: application/json
    {
      "fileId": "tiUxPsiBvLW-JKEVkO93CA",
      "fileExtension": "png"
    }
    
    

    Step 3: Create a Process of the ocrReaders Type

    • Using the file ID you obtained for the source document in Step 2, you can now start the process to OCR the document. This is accomplished by sending a POST request to the Process Manager to start a new OCR process on the OCR Reader Worker.

    Example

    POST http://localhost:3000/v1/processes/ocrReaders
    Content-Type: application/json
    
    {
        "input": {
            "source": {
                "fileId": "XahylGOjtzdGcIdFBH6Wsg"
            },
            "fields": [
                {
                    "area": {
                        "x": 50,
                        "y": 50,
                        "width": 700,
                        "height": 500
                    }
                }
            ],
        "minimumCharacterConfidence": 0,
        "defaultHorizontalResolution": 300
        }
    }
    
    
    200 OK
    Content-Type: application/json
    {
      "processId": "3UFCzaXATPiuHEMws4IWQA",
      "state": "processing",
      "expirationDateTime": "2023-08-03T01:58:10.696Z",
      "input": {
        "source": {
          "fileId": "tiUxPsiBvLW-JKEVkO93CA"
        },
        "fields": [
          {
            "area": {
              "x": 50,
              "y": 50,
              "width": 700,
              "height": 500
            }
          }
        ],
        "minimumCharacterConfidence": 0,
        "defaultHorizontalResolution": 300
      }
    }
    
    

    Step 4: Check Status of the OCR Readers Process

    • The POST request you sent in Step 3 will return immediately before the operation completes. You need to check the status of the process by sending a GET request to the resource you just created.
    • In response to this request, JSON that includes a state property will be returned. When this property is "complete", the JSON response will also include an output property which means you can proceed to the next step. When this property has an "error", the JSON response will include errorCode and errorDetails properties with details about the error.

    Examples

    Successful Response:

    GET http://localhost:3000/v1/processes/ocrReaders/3UFCzaXATPiuHEMws4IWQA
    
    200 OK
    Content-Type: application/json
    {
      "processId": "3UFCzaXATPiuHEMws4IWQA",
      "state": "complete",
      "expirationDateTime": "2023-08-03T01:58:10.000Z",
      "input": {
        "source": {
          "fileId": "tiUxPsiBvLW-JKEVkO93CA"
        },
        "fields": [
          {
            "area": {
              "x": 50,
              "y": 50,
              "width": 700,
              "height": 500
            }
          }
        ],
        "minimumCharacterConfidence": 0,
        "defaultHorizontalResolution": 300
      },
      "output": {
        "fileId": "FdwzTOdlrh8w2gq56YEBUA"
      }
    }
    
    

    Error response when source document is a file format that is not supported by the OCR API:

    GET http://localhost:3000/v1/processes/ocrReaders/3UFCzaXATPiuHEMws4IWQA
    
    200 OK
    Content-Type: application/json
    {
      "input": {
        "source": {
          "fileId": "tiUxPsiBvLW-JKEVkO93CA"
        }
      },
      "processId": "3UFCzaXATPiuHEMws4IWQA",
      "state": "error",
      "errorCode": "UnsupportedFileFormat",
      "errorDetails": {
        "in": "process",
        "at": "input.source.fileId"
      },
      "expirationDateTime": "2023-08-03T01:58:10.000Z"
    }
    
    

    Step 5: Download a File with OCR Results

    • Once the OCR readers process completes successfully, a file with OCR results in JSON format is available for download.
    • A work file ID is made available for each successfully processed document in the output property from the JSON response retrieved in Step 4.
    • See the Work File API for more details about downloading work files.

    Example

    GET http://localhost:18681/PCCIS/V1/WorkFile/FdwzTOdlrh8w2gq56YEBUA
    
    200 OK
    Content-Type: application/json
    {
      "fields": [
        {
          "area": {
            "x": 172,
            "y": 167,
            "width": 425,
            "height": 291
          },
          "confidence": 76,
          "text": "Sample Text",
          "orientation": 0
        }
      ]
    }