MamaSight

MamaSight is a Python package for analyzing images using YOLO object detection and OCR (Optical Character Recognition). It's designed to detect UI elements and text in screenshots and provide annotated visualizations.

Installation

pip install mamasight

Dependencies

MamaSight depends on the following packages:

• torch & torchvision
• ultralytics (v8.3.70)
• numpy (v1.26.4)
• OpenCV
• PaddlePaddle & PaddleOCR (optional for OCR)
• EasyOCR (fallback OCR)
• pandas

Usage

from mamasight import ScreenParser

# Custom box annotation settings (optional)
box_config = {
    'box_overlay_ratio': 3200,  # Base ratio for scaling
    'text_scale': 1.0,          # Scale factor for text
    'text_thickness': 3,        # Thickness of text
    'text_padding': 5,          # Padding around text
    'thickness': 4,             # Thickness of bounding boxes
    'annotation_style': 'simple', # 'simple' or 'colorful'
}

# Initialize parser with custom settings
parser = ScreenParser(box_config=box_config)

# Setup (will auto-detect GPU/CPU if not specified)
parser.setup(yolo_device='cuda', ocr_device='cuda')  # or 'cpu' for CPU

# Analyze image (with or without OCR)
image, detections = parser.analyze('screenshot.png', use_ocr=True)

# Display the annotated image
image.show()

# View detection results as a pandas DataFrame
print(detections)

Features

• Detect UI elements (icons, buttons, etc.) using YOLO
• Optional text detection with OCR
• Customizable annotation styles
• Auto GPU/CPU detection
• Returns both annotated image and structured detection data

License

MIT