Photo Stylization (NPR)
Edge-aware painterly stylization, detail enhancement, and smoothing (OpenCV NPR filters). Three artistic looks, fully in-browser.
Apply non-photorealistic rendering (NPR) to any photo: painterly stylization flattens texture while keeping edges crisp, detail enhance amplifies local contrast, and edge-preserving smoothing gives a clean watercolor-like base. All three are edge-aware filters from OpenCV's photo module, running locally via WebAssembly.
How to Use Photo Stylization (NPR)
- Upload an image: Upload one of the formats listed by the tool (JPG, PNG, WebP) and wait for the local preview.
- Adjust settings and process: Adjust the available controls, then run the selected image processing workflow in your browser.
- Review and download: Review the result, then download it in one of the formats offered by this tool.
Frequently asked questions
What does edge-aware mean?
The filter smooths flat regions aggressively but refuses to average across strong edges, so outlines stay crisp while skin, sky, and walls become flat and painterly.
Which mode should I pick?
Painterly stylization gives the strongest artistic look. Detail enhance is the opposite. It boosts local contrast for a dramatic HDR-like feel. Edge-preserving smooth is subtle and works well as a skin or noise smoother.
How is this different from Cartoonize?
Cartoonize uses a dedicated browser-based cartoon model. These are classical edge-aware filters: deterministic, with device-dependent timing, and tunable with two sliders.
Is my photo uploaded anywhere?
No. The filters are compiled to WebAssembly and run entirely in your browser.