pylightcam 0.1.0

极轻量 Windows 摄像头拍照库 — 无需 OpenCV，即装即用

pylightcam

Ultra-lightweight Windows camera capture library — no OpenCV required, works out of the box.

Installation

pip install pylightcam

Quick Start

import pylightcam as plc

cap = plc.VideoCapture(0)
ret, img = cap.read()

if ret:
    # img is a numpy.ndarray (H×W×3, RGB)
    # Save with Pillow
    from PIL import Image
    Image.fromarray(img).save("photo.jpg")

cap.release()

Or use Pillow directly:

from PIL import Image

cap = plc.VideoCapture(0)
ret, pil_img = cap.read()  # pil_img is PIL.Image
if ret:
    pil_img.save("photo.jpg")

Pure JPEG bytes (zero dependencies):

cap = plc.VideoCapture(0)
ret, jpg_bytes = cap.read_jpg()
if ret:
    with open("photo.jpg", "wb") as f:
        f.write(jpg_bytes)

API

plc.get_device_count()

Returns the number of cameras on the system.

plc.VideoCapture(device_index=0, width=0, height=0)

Opens a camera. Supports with statement.

Return type	Method	Dependencies
numpy.ndarray (RGB)	.read()	numpy + pillow
PIL.Image	.read()	pillow
bytes (JPEG)	.read_jpg()	none

.release()

Close the camera.

Comparison with cv2.VideoCapture

Feature	cv2	pylightcam
Package size	~500 MB	~50 KB
Dependencies	OpenCV + ffmpeg + ...	none
Color format	BGR	RGB
Capture speed	fast	moderate