SafeScanEngine 0.2.0

A package to identify dangerous non baby-proof items.

SafeScanEngine

Overview

SafeScanEngine is a Python Library designed for identifying dangerous (non baby-proof) from a list of images. The engine was design to be an interface for more than one image processing solvers. For the time being, YoloV8 was implemented

Key Features

• Can identify both lethal and non-lethal objects.

Lethal:

• Electrical outlets
• Plastic bags (to be implemented)
• Cables/cords of any kind (to be implemented)

Non-lethal:

• Sharp corners (to be implemented)

Requirements

• Python 3.7 or higher
• Pillow

Installation

To install, run:

pip install SafeScanEngine

Usage

from safe_scan_engine import YoloV8Resolver

model = YoloV8Resolver(images=["img_1.jpg", "img_no_danger.jpg", "img_2.jpg"])
results = model.return_detections()

# Process results list
for result in results:
    boxes = result["boxes"]  # Boxes as list for bbox outputs in json serializable form
    original_shape = result["original_shape"]  # Input image dimensions
    labels = result["names"]  # Labels detected (e.g. ElectricalOutlet) 
    result_image_with_label = result["encoded_image"]  # Result image (with boxes) encoded in base64
    was_danger_detected = result["danger_found"]  # Boolean always available even if nothing was detected (e.g. False)

License

MIT License. See LICENSE for details.

Contact

Reach out at [e-mail] for support or inquiries.