ᛋ sowilo

Differentiable computer vision for OCaml

Differentiable computer vision on Rune tensors.

Sowilo provides image processing operations expressed purely through Rune tensor operations. Filters, edge detectors, morphological operations, color transforms, and geometric transforms -- all compatible with Rune.grad and Rune.vmap.

Image Conventions

Images are Nx.t tensors with channels-last layout:

  • Single image: [H; W; C] (height, width, channels)
  • Batch: [N; H; W; C] (batch, height, width, channels)
  • Grayscale: C = 1, RGB: C = 3, RGBA: C = 4

Operations expect float32 tensors with values in [0, 1]. Use to_float to convert from uint8 and to_uint8 to convert back.

What's Included

  • Type conversion: to_float, to_uint8, normalize, threshold
  • Color: to_grayscale, rgb_to_hsv, hsv_to_rgb, brightness, contrast, saturation, hue, gamma, invert
  • Geometric transforms: resize, crop, center_crop, hflip, vflip, rotate90, pad
  • Spatial filters: gaussian_blur, box_blur, median_blur, filter2d, unsharp_mask
  • Morphology: structuring_element, erode, dilate, opening, closing, morphological_gradient
  • Edge detection: sobel, scharr, laplacian, canny

Quick Start

open Sowilo

let () =
  (* Load image and convert to float32 [0, 1] *)
  let img = Nx_io.load_image "photo.png" |> to_float in

  (* Process: grayscale, blur, edge detection *)
  let edges =
    img
    |> to_grayscale
    |> gaussian_blur ~sigma:1.0
    |> canny ~low:0.2 ~high:0.6
  in

  (* Save result *)
  Nx_io.save_image (to_uint8 edges) "edges.png"

Learn More