Read & Extract Text from Images via Python API

Getting Started with Aspose.OCR for Python via .NET

The recommend way to install Aspose.OCR for Python via .NET is using pip. Please use the following command for a smooth installation.

 pip install aspose-ocr-python-net

OCR Operations with High Accuracy via Python

Aspose.OCR for Python via .NET is engineered for high precision and accuracy. The library incorporates advanced machine learning models that improve text extraction accuracy, even with skewed or low-resolution images. This feature makes it suitable for applications that require reliable text recognition, such as automated data extraction from scanned forms or documents. The following code snippet showcases a simple implementation where an image is loaded, processed, and its recognized text is displayed.

# Initialize OCR engine
recognitionEngine = AsposeOcr()

# Add image to batch
input = OcrInput(InputType.SINGLE_IMAGE)
input.add("sample.png")

# Extract text from image
result = recognitionEngine.recognize(input)
# Display the recognition result
print(result[0].recognition_text)

Image Preprocessing Capabilities

Aspose.OCR for Python via .NET library provides powerful image preprocessing features inside Python applications. The features enhance OCR accuracy, such as skew correction, noise removal, and image normalization. These preprocessing steps are crucial when working with images that may have been scanned under suboptimal conditions. The following example demonstrates how developers can perform the skew correction operation which ensures that text is accurately detected, even if the image is slightly tilted or not perfectly aligned.

from aspose.ocr import OcrEngine, SkewCorrection

ocr_engine = OcrEngine()

# Enable skew correction
ocr_engine.set_skew_correction(SkewCorrection.True)

ocr_engine.image = "skewed_image.png"
recognized_text = ocr_engine.get_text()
print("Corrected and Recognized Text:", recognized_text)

Handwritten Text Recognition via Python

Aspose.OCR for Python via .NET is very easy to handle and can recognize both printed and handwritten text with just a couple of lines of Python code. This features is very useful for applications like digitizing handwritten notes or signatures. The software’s ability to interpret various styles of text enhances its utility in sectors like education and legal documentation. Here is an example that shows how to perform Handwritten Text Recognition using Python code.

api = ocr.AsposeOcr()

''' add filters if you need '''
filters = ocr.models.preprocessingfilters.PreprocessingFilter()
#filters.add(ocr.models.preprocessingfilters.PreprocessingFilter.contrast_correction_filter())

''' initialize image collection and put images into it '''
input = ocr.OcrInput(ocr.InputType.SINGLE_IMAGE, filters)
input.add("Data\\OCR\\handwritten.jpg")

''' change recognition options if you need '''
settings = ocr.RecognitionSettings()
settings.detect_areas_mode=ocr.DetectAreasMode.PHOTO

''' run recognition '''
res = api.recognize_handwritten_text(input)
print(res[0].recognition_text)

OCR Custom Image Regions in Python Apps

Aspose.OCR for Python via .NET has provided complete support for recognizing text from a particular area of an image inside Python applications. Software developers can specify specific regions within an image for OCR, which is helpful in scenarios where only a portion of the image contains relevant text. Developers can define custom regions for OCR on an image, set recognition modes, and adjust other parameters to optimize the OCR process based on specific application requirements. Here is an example that shows how software developers can recognize a single line of text with just a couple of lines of Python code.

api = AsposeOcr()

# Create OcrInput and add images
input = OcrInput(InputType.SINGLE_IMAGE)
input.add("sample_line.png")

# recognize without regions detection
settings = RecognitionSettings()
settings.recognize_single_line = True

result = api.recognize(input, settings)

print(result[0].recognition_text)