Convert a scanned image into words with location

v1.0.0

Converts an uploaded image in common formats such as JPEG, PNG into lines/text with location information and other metdata via Optical Character Recognition. This API is intended to be run on scanned documents. If you want to OCR photos (e.g. taken with a smart phone camera), be sure to use the photo/toText API instead, as it is designed to unskew the image first. Note: for free tier API keys, it is required to add a credit card to your account for security reasons, to use the free tier key with this API.

Input

{
  "type": "object",
  "properties": {
    "language": {
      "type": "string"
    },
    "preprocessing": {
      "type": "string"
    }
  },
  "additionalProperties": false
}

Output

{
  "type": "object",
  "description": "Result of an image to lines-with-location OCR operation",
  "properties": {
    "Lines": {
      "type": "array",
      "description": "Words in the image",
      "items": {
        "type": "object",
        "description": "A contiguous line of text in an OCR document",
        "properties": {
          "LineText": {
            "type": "string",
            "description": "Text of the line"
          },
          "Words": {
            "type": "array",
            "description": "Word objects in the line",
            "items": {
              "type": "object",
              "description": "A single word in an OCR document",
              "properties": {
                "BlockNumber": {
                  "type": "integer",
                  "format": "int32",
                  "description": "Index of the containing block"
                },
                "ConfidenceLevel": {
                  "type": "number",
                  "format": "double",
                  "description": "Confidence level of the machine learning result; possible values are 0.0 (lowest accuracy) - 1.0 (highest accuracy)"
                },
                "Height": {
                  "type": "integer",
                  "format": "int32",
                  "description": "Height of the word in pixels"
                },
                "LineNumber": {
                  "type": "integer",
                  "format": "int32",
                  "description": "Line number of the word"
                },
                "PageNumber": {
                  "type": "integer",
                  "format": "int32",
                  "description": "Index of the containing page"
                },
                "ParagraphNumber": {
                  "type": "integer",
                  "format": "int32",
                  "description": "Index of the containing paragraph"
                },
                "Width": {
                  "type": "integer",
                  "format": "int32",
                  "description": "Width of the word in pixels"
                },
                "WordNumber": {
                  "type": "integer",
                  "format": "int32",
                  "description": "Index of the word in the line"
                },
                "WordText": {
                  "type": "string",
                  "description": "Text of the word"
                },
                "XLeft": {
                  "type": "integer",
                  "format": "int32",
                  "description": "X location of the left edge of the word in pixels"
                },
                "YTop": {
                  "type": "integer",
                  "format": "int32",
                  "description": "Y location of the top edge of the word in pixels"
                }
              }
            }
          }
        }
      }
    },
    "Successful": {
      "type": "boolean",
      "description": "True if successful, false otherwise"
    }
  }
}