Invoice OCR Python

The Python OCR SDK supports the Invoice API.

Using the sample below, we are going to illustrate how to extract the data that we want using the OCR SDK.
Invoice sample

Quick-Start

from mindee import Client, PredictResponse, product

# Init a new client
mindee_client = Client(api_key="my-api-key")

# Load a file from disk
input_doc = mindee_client.source_from_path("/path/to/the/file.ext")

# Load a file from disk and parse it.
# The endpoint name must be specified since it cannot be determined from the class.
result: PredictResponse = mindee_client.parse(product.InvoiceV4, input_doc)

# Print a summary of the API result
print(result.document)

# Print the document-level summary
# print(result.document.inference.prediction)

Output (RST):

########
Document
########
:Mindee ID: 80f2328c-58a5-486a-9599-eb2d738680f0
:Filename: default_sample.jpg

Inference
#########
:Product: mindee/invoices v4.4
:Rotation applied: Yes

Prediction
==========
:Locale: en; en; CAD;
:Invoice Number: 14
:Reference Numbers: AD29094
:Purchase Date: 2018-09-25
:Due Date: 2018-09-25
:Total Net:
:Total Amount: 2608.20
:Total Tax: 193.20
:Taxes:
  +---------------+--------+----------+---------------+
  | Base          | Code   | Rate (%) | Amount        |
  +===============+========+==========+===============+
  |               |        | 8.00     | 193.20        |
  +---------------+--------+----------+---------------+
:Supplier Payment Details:
:Supplier Name: TURNPIKE DESIGNS
:Supplier Company Registrations:
:Supplier Address: 156 University Ave, Toronto ON, Canada M5H 2H7
:Customer Name: JIRO DOI
:Customer Company Registrations:
:Customer Address: 1954 Bloor Street West Toronto, ON, M6P 3K9 Canada
:Document Type: INVOICE
:Line Items:
  +--------------------------------------+--------------+----------+------------+--------------+--------------+------------+
  | Description                          | Product code | Quantity | Tax Amount | Tax Rate (%) | Total Amount | Unit Price |
  +======================================+==============+==========+============+==============+==============+============+
  | Platinum web hosting package Down... |              | 1.00     |            |              | 65.00        | 65.00      |
  +--------------------------------------+--------------+----------+------------+--------------+--------------+------------+
  | 2 page website design Includes ba... |              | 3.00     |            |              | 2100.00      | 2100.00    |
  +--------------------------------------+--------------+----------+------------+--------------+--------------+------------+
  | Mobile designs Includes responsiv... |              | 1.00     |            |              | 250.00       | 250.00     |
  +--------------------------------------+--------------+----------+------------+--------------+--------------+------------+

Page Predictions
================

Page 0
------
:Locale: en; en; CAD;
:Invoice Number: 14
:Reference Numbers: AD29094
:Purchase Date: 2018-09-25
:Due Date: 2018-09-25
:Total Net:
:Total Amount: 2608.20
:Total Tax: 193.20
:Taxes:
  +---------------+--------+----------+---------------+
  | Base          | Code   | Rate (%) | Amount        |
  +===============+========+==========+===============+
  |               |        | 8.00     | 193.20        |
  +---------------+--------+----------+---------------+
:Supplier Payment Details:
:Supplier Name: TURNPIKE DESIGNS
:Supplier Company Registrations:
:Supplier Address: 156 University Ave, Toronto ON, Canada M5H 2H7
:Customer Name: JIRO DOI
:Customer Company Registrations:
:Customer Address: 1954 Bloor Street West Toronto, ON, M6P 3K9 Canada
:Document Type: INVOICE
:Line Items:
  +--------------------------------------+--------------+----------+------------+--------------+--------------+------------+
  | Description                          | Product code | Quantity | Tax Amount | Tax Rate (%) | Total Amount | Unit Price |
  +======================================+==============+==========+============+==============+==============+============+
  | Platinum web hosting package Down... |              | 1.00     |            |              | 65.00        | 65.00      |
  +--------------------------------------+--------------+----------+------------+--------------+--------------+------------+
  | 2 page website design Includes ba... |              | 3.00     |            |              | 2100.00      | 2100.00    |
  +--------------------------------------+--------------+----------+------------+--------------+--------------+------------+
  | Mobile designs Includes responsiv... |              | 1.00     |            |              | 250.00       | 250.00     |
  +--------------------------------------+--------------+----------+------------+--------------+--------------+------------+

Field Types

Standard Fields

These fields are generic and used in several products.

BasicField

Each prediction object contains a set of fields that inherit from the generic BaseField class.
A typical BaseField object will have the following attributes:

  • value (Union[float, str]): corresponds to the field value. Can be None if no value was extracted.
  • confidence (float): the confidence score of the field prediction.
  • bounding_box ([Point, Point, Point, Point]): contains exactly 4 relative vertices (points) coordinates of a right rectangle containing the field in the document.
  • polygon (List[Point]): contains the relative vertices coordinates (Point) of a polygon containing the field in the image.
  • page_id (int): the ID of the page, is None when at document-level.
  • reconstructed (bool): indicates whether an object was reconstructed (not extracted as the API gave it).

Note: A Point simply refers to a List of two numbers ([float, float]).

Aside from the previous attributes, all basic fields have access to a custom __str__ method that can be used to print their value as a string.

AmountField

The amount field AmountField only has one constraint: its value is an Optional[float].

ClassificationField

The classification field ClassificationField does not implement all the basic BaseField attributes. It only implements value, confidence and page_id.

Note: a classification field's value is always a str`.

CompanyRegistrationField

Aside from the basic BaseField attributes, the company registration field CompanyRegistrationField also implements the following:

  • type (str): the type of company.

DateField

Aside from the basic BaseField attributes, the date field DateField also implements the following:

  • date_object (Date): an accessible representation of the value as a python object. Can be None.

LocaleField

The locale field LocaleField only implements the value, confidence and page_id base BaseField attributes, but it comes with its own:

  • language (str): ISO 639-1 language code (e.g.: en for English). Can be None.
  • country (str): ISO 3166-1 alpha-2 or ISO 3166-1 alpha-3 code for countries (e.g.: GRB or GB for "Great Britain"). Can be None.
  • currency (str): ISO 4217 code for currencies (e.g.: USD for "US Dollars"). Can be None.

PaymentDetailsField

Aside from the basic BaseField attributes, the payment details field PaymentDetailsField also implements the following:

  • account_number (str): number of an account, expressed as a string. Can be None.
  • iban (str): International Bank Account Number. Can be None.
  • routing_number (str): routing number of an account. Can be None.
  • swift (str): the account holder's bank's SWIFT Business Identifier Code (BIC). Can be None.

StringField

The text field StringField only has one constraint: its value is an Optional[str].

TaxesField

Tax

Aside from the basic BaseField attributes, the tax field TaxField also implements the following:

  • rate (float): the tax rate applied to an item expressed as a percentage. Can be None.
  • code (str): tax code (or equivalent, depending on the origin of the document). Can be None.
  • base (float): base amount used for the tax. Can be None.

Note: currently TaxField is not used on its own, and is accessed through a parent Taxes object, a list-like structure.

Taxes (Array)

The Taxes field represents a list-like collection of TaxField objects. As it is the representation of several objects, it has access to a custom __str__ method that can render a TaxField object as a table line.

Specific Fields

Fields which are specific to this product; they are not used in any other product.

Line Items Field

List of line item details.

A InvoiceV4LineItem implements the following attributes:

  • description (str): The item description.
  • product_code (str): The product code referring to the item.
  • quantity (float): The item quantity
  • tax_amount (float): The item tax amount.
  • tax_rate (float): The item tax rate in percentage.
  • total_amount (float): The item total amount.
  • unit_price (float): The item unit price.

Attributes

The following fields are extracted for Invoice V4:

Customer Address

customer_address (StringField): The address of the customer.

print(result.document.inference.prediction.customer_address.value)

Customer Company Registrations

customer_company_registrations (List[CompanyRegistrationField]): List of company registrations associated to the customer.

for customer_company_registrations_elem in result.document.inference.prediction.customer_company_registrations:
    print(customer_company_registrations_elem.value)

Customer Name

customer_name (StringField): The name of the customer or client.

print(result.document.inference.prediction.customer_name.value)

Purchase Date

date (DateField): The date the purchase was made.

print(result.document.inference.prediction.date.value)

Document Type

document_type (ClassificationField): One of: 'INVOICE', 'CREDIT NOTE'.

print(result.document.inference.prediction.document_type.value)

Due Date

due_date (DateField): The date on which the payment is due.

print(result.document.inference.prediction.due_date.value)

Invoice Number

invoice_number (StringField): The invoice number or identifier.

print(result.document.inference.prediction.invoice_number.value)

Line Items

line_items (List[InvoiceV4LineItem]): List of line item details.

for line_items_elem in result.document.inference.prediction.line_items:
    print(line_items_elem)

Locale

locale (LocaleField): The locale detected on the document.

print(result.document.inference.prediction.locale.value)

Reference Numbers

reference_numbers (List[StringField]): List of Reference numbers, including PO number.

for reference_numbers_elem in result.document.inference.prediction.reference_numbers:
    print(reference_numbers_elem.value)

Supplier Address

supplier_address (StringField): The address of the supplier or merchant.

print(result.document.inference.prediction.supplier_address.value)

Supplier Company Registrations

supplier_company_registrations (List[CompanyRegistrationField]): List of company registrations associated to the supplier.

for supplier_company_registrations_elem in result.document.inference.prediction.supplier_company_registrations:
    print(supplier_company_registrations_elem.value)

Supplier Name

supplier_name (StringField): The name of the supplier or merchant.

print(result.document.inference.prediction.supplier_name.value)

Supplier Payment Details

supplier_payment_details (List[PaymentDetailsField]): List of payment details associated to the supplier.

for supplier_payment_details_elem in result.document.inference.prediction.supplier_payment_details:
    print(supplier_payment_details_elem.value)

Taxes

taxes (List[TaxField]): List of tax line details.

for taxes_elem in result.document.inference.prediction.taxes:
    print(taxes_elem.polygon)

Total Amount

total_amount (AmountField): The total amount paid: includes taxes, tips, fees, and other charges.

print(result.document.inference.prediction.total_amount.value)

Total Net

total_net (AmountField): The net amount paid: does not include taxes, fees, and discounts.

print(result.document.inference.prediction.total_net.value)

Total Tax

total_tax (AmountField): The total tax: includes all the taxes paid for this invoice.

print(result.document.inference.prediction.total_tax.value)

Questions?

Join our Slack