Ctrl+K

Easy OCR Model

Easy OCR Model#

download_image Function#

download_image(image_url: str) → Image#

Downloads an image from a URL into a format processable for Optical Character Recognition (OCR).

Parameters:

image_url (str) – The URL of the image.

Returns:

A Pillow Image object created from the downloaded image.

Retrieves an image from the specified URL and converts it into a Pillow Image object for further processing.

draw_boxes Function#

draw_boxes(image: Image, bounds, color='yellow', width=2) → Image#

Draws bounding boxes around text found in the image.

Parameters:

  • image (Image) – The Pillow Image object.

  • bounds – An iterator containing the bounding box coordinates.

  • color (str) – The color of the bounding boxes (default is ‘yellow’).

  • width (int) – The width of the bounding box lines (default is 2).

Returns:

The updated Pillow Image object with bounding boxes drawn around the text.

Draws bounding boxes around text regions detected in the image. It is primarily used for visualization purposes.

inference Function#

inference(img_source, lang='en') → list[str]#

Reads text from an image and returns the detected text.

Parameters:

  • img_source (str) – The source of the image (can be a URL or a file path).

  • lang (str) – The language used for text detection (default is ‘en’ for English).

Returns:

A list of strings containing the detected text.

Reads text from the input image using the EasyOCR library. It supports both local image files and URLs. Text detection is performed based on the specified language, with an optional parameter to adjust confidence filtering.

Example usage:#

from ocr_module import download_image, inference

image_url = "https://example.com/image.jpg"
img = download_image(image_url)
text = inference(img)
print("Detected Text:", text)
Contents