iris-cv 0.0.0

A fast computer vision library framework in Rust.
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34

// Demonstrates Canny edge detection on a real image.
// Loads the test pattern and applies Canny with two threshold levels.

use burn::backend::wgpu::{Wgpu, WgpuDevice};
use iris::prelude::*;

fn main() -> Result<()> {
    type Backend = Wgpu;
    let device = WgpuDevice::default();

    println!(
        "Using compute backend: {}",
        BurnUtils::backend_name::<Backend>()
    );

    // Load the test pattern image
    let image: Image<Backend> = Image::open("assets/images/test_pattern.png", &device)?;
    println!(
        "Loaded image: {}x{} ({} channels)",
        image.width(),
        image.height(),
        image.channels()
    );

    // Perform Canny edge detection
    let edges = image.canny(0.1, 0.4)?;
    println!("Edges shape: {:?}", edges.shape());

    // Save the result
    edges.save("output_canny.png")?;
    println!("Saved Canny edges to 'output_canny.png'");

    Ok(())
}