OCR - AWS Rekognition Node for TextDetection

Feature Requests Nodes
1

Hello,
I need some OCR functionality and wanted to use AWS Rekognition.

At the moment the AWS Rekognition Node is not able to use the Rekognition API for Detect Text. Can you please add the option or is there any other node, which provides those functionionality?

Thanks in advance.

2

@ixidion Got added. If you can run it locally and provide feedback would be awesome.

https://github.com/n8n-io/n8n/pull/1417/commits/d221a149824046945f55e861ab56c2a4a38bd0c9

Test workflow:

{
  "nodes": [
    {
      "parameters": {},
      "name": "Start",
      "type": "n8n-nodes-base.start",
      "typeVersion": 1,
      "position": [
        250,
        300
      ]
    },
    {
      "parameters": {
        "type": "detectText",
        "binaryData": true,
        "additionalFields": {}
      },
      "name": "AWS Rekognition",
      "type": "n8n-nodes-base.awsRekognition",
      "typeVersion": 1,
      "position": [
        870,
        300
      ],
      "credentials": {
        "aws": "n8n"
      }
    },
    {
      "parameters": {
        "url": "https://dhei5unw3vrsx.cloudfront.net/images/coffee_monday_resized.jpg",
        "responseFormat": "file",
        "options": {}
      },
      "name": "HTTP Request",
      "type": "n8n-nodes-base.httpRequest",
      "typeVersion": 1,
      "position": [
        550,
        300
      ]
    }
  ],
  "connections": {
    "Start": {
      "main": [
        [
          {
            "node": "HTTP Request",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "HTTP Request": {
      "main": [
        [
          {
            "node": "AWS Rekognition",
            "type": "main",
            "index": 0
          }
        ]
      ]
    }
  }
}
2 Likes
3

Hi @RicardoE105 ,

thanks for the implementation. The feature does what I need and detects the text in the image.

Well, have tested the options “min confidence” / “Regions of Interest”. They seem to have no function at all. I don’t need them, but maybe there is sth broken.

e.g. I set a region without text and the Text is still detected.
min is 100 and still detects the 98% confidecne.

4

Thanks for letting us know. Will investigate.

6

Found the issue and got fixed. @ixidion

2 Likes
7

Great. When will this be approx. in the Docker Image?

1 Like
8

Probably on Monday