This is an image smoothing algorithm inspired by JPEG Quant Smooth. This is called gravity blur because the pixel values gravitate towards each other, the force of this gravity is inversely proportional to the distance between them.

This algorithm preserves sharp edges if the distance between pixel values on the sides of the edges is greater than range. With more iterations, it is not only smoothes the image, but can also make it look like CG art or poster.

The original project page is here.

Web version available here. Images are processed locally on your computer.

• Click the "Load" button or drag-n-drop image into the browser window. After processing is complete, you can save the result by clicking the "Save" button. You can edit the filename field before saving to change the image format (jpeg or png, JPEG images are saved at 100% quality). The options field is passed to the wasm code when you initiate a processing by loading a file.

gravblur [options] input.[png|jpg|jpeg] output.[png|jpg|jpeg]

-r, --range f Gravity range (0-100)
-s, --sharp f Sharpness (0-100)
-n, --niter n Number of iterations (default is 3)
-R, --rgb Process in RGB (default)
-Y, --yuv Process in YUV
-S, --separate Separate color components
-a, --alpha Use alpha channel
-i, --info n Print gravblur debug output: 0 - silent, 3 - all (default)

• Images 3x zoomed.

If your system have libjpeg and libpng development packages installed, just type make. Or you can specify their location with LIBS variable to make.