Eye

Eye is a cross platform camera capture and control library written in native Rust. It features multiple platform backends, such as v4l2 for Linux. Buffers are captured by accessing 'streams'. The stream concept is used to facilitate additional features such as colorspace conversion.

A backend is also called 'HAL' aka hardware abstraction layer in eye-rs. An OS may support multiple HALs at runtime, depending on which HALs were selected at compile time. Capture devices are then identified via their URI so you can choose the HAL which best fits your needs.

Eye is a very young library and its API is subject to change (as denoted by the 0.x.x version number). We follow the semver approach, meaning each new feature will bump the minor version by one.

Goals

Eye strives to provide a common feature set on all platforms. Some devices, mostly more expensive ones, will always offer more features than others though. Eye shall expose a dynamic featureset API which can be queried at runtime so device parameters can be configured accordingly.

Common Features

• Transparent pixel format conversion

OS Feature Matrix

There are various HAL specific properties. For example, the v4l2 HAL on Linux supports zero-copy capture (as far as userspace is concerned - the kernel driver may still perform a copy). Those will be enumerated here in the future.

Usage

Below you can find a quick example usage of this crate. It introduces the basics necessary for image capturing.

use eye_hal::PlatformContext;
use eye_hal::traits::{Context, Device, Stream};

fn main() -> Result<()> {
    // Create a context
    let ctx = PlatformContext::default();

    // Query for available devices.
    let devices = ctx.query_devices()?;

    // First, we need a capture device to read images from. For this example, let's just choose
    // whatever device is first in the list.
    let dev = ctx.open_device(&devices[0])?;

    // Query for available streams and just choose the first one.
    let streams = dev.query_streams()?;
    let stream_desc = streams[0].clone();
    println!("Stream: {:?}", stream_desc);

    // Since we want to capture images, we need to access the native image stream of the device.
    // The backend will internally select a suitable implementation for the platform stream. On
    // Linux for example, most devices support memory-mapped buffers.
    let mut stream = dev.start_stream(&stream_desc)?;

    // Here we create a loop and just capture images as long as the device produces them. Normally,
    // this loop will run forever unless we unplug the camera or exit the program.
    loop {
        let frame = stream
            .next()
            .expect("Stream is dead")
            .expect("Failed to capture frame");
    }
}

Have a look at the provided examples for more sample applications.