Where OpenCV shines
OpenCV is the standard open-source computer vision library, licensed under Apache 2.0. It covers detection, tracking, camera calibration, feature matching, video I/O, and a large set of image processing functions, and its Python bindings work directly with NumPy arrays. For real-time video or anything in the computer vision and machine learning world, it's the natural choice.
Do it with BrushCue
BrushCue is built for a different job: creative image editing. Its effects come ready-made, run on the GPU, and blend in linear light with color management. For a simple operation, the two look similar:
OpenCV
import cv2
image = cv2.imread("photo.jpg")
blurred = cv2.GaussianBlur(image, (0, 0), 8)
cv2.imwrite("photo-blurred.png", blurred)BrushCue (Python)
import brushcue as bc
ctx = bc.Context()
image = bc.Composition.load("photo.jpg")
# Gaussian blur with a sigma of 8 pixels
edited = image.gaussian_blur(8.0)
result = edited.execute(ctx)
with open("photo-edited.png", "wb") as f:
f.write(result.to_image_bytes(ctx))For a single blur, OpenCV is just as short, and it's a great fit when the result feeds into vision code. The difference shows with creative effects: BrushCue ships halftones, swirls, film grain, and color grades as single calls, where in OpenCV you'd write the pixel math yourself.
BrushCue Script (bcue)
image = Composition::from_path("photo.jpg")
edited = image.gaussian_blur(8.0)
Bytes::save_to_path(Image::to_bytes(Composition::render_to_image(edited)), "photo-edited.png")



