qimage-hwc-bgrx-8888-ndarray-interop 0.1.0a0

Zero-copy conversion between Qt QImage (in `Format_RGB32`/`Format_ARGB32`) and NumPy ndarray (in HWC BGRX 8888 format).

qimage-hwc-bgrx-8888-ndarray-interop

Zero-copy conversion between Qt QImage (in Format_RGB32/Format_ARGB32) and NumPy ndarray (in HWC BGRX 8888 format).

Features

• Zero-copy: Memory is shared between objects - no pointless data copying!
• Cross-binding: Supports PyQt6, PyQt5, PyQt4, PySide6, PySide2, or PySide.
• Safety: Explicit checks for packing and format - no surprises.

Installation

Install your preferred Qt binding, then run:

pip install qimage-hwc-bgrx-8888-ndarray-interop

Usage

# coding=utf-8
from qimage_hwc_bgrx_8888_ndarray_interop import (
    qimage_to_hwc_bgrx_8888_ndarray_view,
    qimage_to_hwc_bgrx_8888_ndarray,
    hwc_bgrx_8888_ndarray_to_qimage_view,
    hwc_bgrx_8888_ndarray_to_qimage,
)

# Convert a QImage to numpy.ndarray (zero-copy view):
arr_view = qimage_to_hwc_bgrx_8888_ndarray_view(qimage)

# Convert a QImage to a copy of image data:
arr = qimage_to_hwc_bgrx_8888_ndarray(qimage)

# Convert a contiguous (H, W, 4) uint8 ndarray back to a QImage (zero-copy view):
qimg_view = hwc_bgrx_8888_ndarray_to_qimage_view(arr)

# Convert ndarray to a QImage (copies data):
qimg_copy = hwc_bgrx_8888_ndarray_to_qimage(arr)

Safety & Caveats

• Endianness: Only little-endian systems are supported (typical modern computers).
  • On unsupported (big-endian) systems, RuntimeError is raised at import time.
• Tightly packed images only: QImage must have bytesPerLine == width * 4 for zero-copy view.
• Memory lifetime matters:
  • Do not let the QImage or ndarray be garbage collected while the other exists and references its buffer.
  • If in doubt, use the copy() variants.

Contributing

Contributions are welcome! Please submit pull requests or open issues on the GitHub repository.

License

This project is licensed under the MIT License.