No difference in output with OCREngine 1 / 2, and isTable true / false

OCR.space Free OCR API
1

I’ve tried all combinations of params listed in the title and the parsedText is identical for all.

For testing I’m using postman. So simple http requests.
Also, I’m using the pro endpoint as a paid pro customer.

Here’s the image.

1500×1500 234 KB

And here's the response 
{
    "ParsedResults": [
        {
            "TextOverlay": {
                "Lines": [],
                "HasOverlay": false,
                "Message": "Text overlay is not provided as it is not requested"
            },
            "TextOrientation": "0",
            "FileParseExitCode": 1,
            "ParsedText": "Supplement Facts\r\nSuggested Serving Size: 2 Capsule\r\nServings Per Container: 60\r\nAmount Per Serving\r\n% Daily Value\r\nMagnesium\r\n266 mg\r\n(from 1746 mg of magnesium citrate)\r\nZinc\r\n(from 54 mg of zinc picolinate)\r\n10 mg\r\nOther Ingredients: Vegetable Capsule (Hypromellose),\r\nOrganic Rice (Hull)\r\nSuggested Adult Use: To supplement a healthy diet take\r\n1-2 capsules daily, in the evening.\r\nCaution: Please consult your healthcare provider before\r\ntaking this or any supplement if you are pregnant, breast-\r\nfeeding, and/or taking any OTC or prescription medications.\r\nStore in a cool, dry place. Keep out of reach of children and\r\npets. Do not use if safety seal is broken or missing.\r\nThird\r\nParty\r\nCertified\r\nFor Purity & Accuracy\r\n",
            "ErrorMessage": "",
            "ErrorDetails": ""
        }
    ],
    "OCRExitCode": 1,
    "IsErroredOnProcessing": false,
    "ProcessingTimeInMilliseconds": "656",
    "SearchablePDFURL": "Searchable PDF not generated as it was not requested."
}

Is this normal? At a minimum I would expect isTable to produce different output.
Also, there’s no indication in the ParsedResults response indicating which parameters it used.

Thanks!

2

Can you please post a screenshot of your Postman OCR API call?

You are correct: There should be a difference between using OCR Engine 1/2 and table mode on/off. Something might be wrong in the API call.

4

That is correct. The used parameters are not included in the API response. But see my questions above. The OCR result should be different.

5

Here’s an HTTP request provided by Postmans code snippet section.

POST /parse/image?OCREngine=2&isTable=true HTTP/1.1
Host: apipro1.ocr.space
apiKey: xxxxxxxxxxx
Content-Length: 190
Content-Type: multipart/form-data; boundary=----WebKitFormBoundary7MA4YWxkTrZu0gW

------WebKitFormBoundary7MA4YWxkTrZu0gW
Content-Disposition: form-data; name="url"

https://m.media-amazon.com/images/I/81kBjaK8iSL._AC_SL1500_.jpg
------WebKitFormBoundary7MA4YWxkTrZu0gW--
6

image
image1015×404 66.8 KB

7

/parse/image?OCREngine=2&isTable=true HTTP/1.1

That looks wrong. To be sure, can you please post a screenshot of the call inside the postman app?

Something like this: https://ocr.space/Content/Images/postman-url-ocr.webp

8

Got it. Thanks!

I was passing the options in as Params, but they need to be part of the form-data body.
Your screenshot provided clarification.

:beers: