rawkit

A Go library for reading and processing RAW image files. Wraps LibRaw via cgo with fallback support for standard image formats (JPEG, PNG).

go get github.com/tlaceby/rawkit@latest

Supported Formats

RAW formats: ARW (Sony), CR2/CR3 (Canon), NEF (Nikon), DNG (Adobe), ORF (Olympus), RAF (Fujifilm), RW2 (Panasonic)

Standard formats: JPEG, PNG

Requirements

LibRaw must be installed on your system.

# macOS
brew install libraw

# Ubuntu/Debian
sudo apt install libraw-dev

# Fedora
sudo dnf install LibRaw-devel

Packages

Package	Description
core	Image reading, metadata extraction, pixel data
histogram	Histogram generation and analysis
processing	Image adjustments and transformations

For detailed API documentation, see pkg.go.dev/github.com/tlaceby/rawkit or run:

go doc github.com/tlaceby/rawkit/core
go doc github.com/tlaceby/rawkit/histogram
go doc github.com/tlaceby/rawkit/processing

Examples

Read an Image

package main

import (
    "fmt"
    "log"

    "github.com/tlaceby/rawkit/core"
)

func main() {
    img, err := core.ReadAll("/path/to/photo.arw")
    if err != nil {
        log.Fatal(err)
    }

    fmt.Printf("Dimensions: %dx%d\n", img.Data.Width, img.Data.Height)
    fmt.Printf("Bit Depth: %d\n", img.Data.BitDepth)

    if img.Meta != nil {
        fmt.Printf("Camera: %s %s\n", img.Meta.CameraMake, img.Meta.CameraModel)
        fmt.Printf("ISO: %d, f/%.1f, %.4fs\n", img.Meta.ISO, img.Meta.Aperture, img.Meta.SS)
    }
}

Metadata Only (Fast)

meta, err := core.ReadRAWMetadata("/path/to/photo.arw")
if err != nil {
    log.Fatal(err)
}

fmt.Printf("Camera: %s %s\n", meta.CameraMake, meta.CameraModel)

Pixel Data Only

data, err := core.ReadImageData("/path/to/photo.arw")
if err != nil {
    log.Fatal(err)
}

// Access pixel at (100, 200)
r, g, b := data.Pixel(100, 200)

Pixel Data Format

All pixel data is []uint16 in RGB layout:

// Using helper methods
r, g, b := data.Pixel(x, y)

// Direct buffer access
idx := data.Index(x, y)
r, g, b := data.Data[idx], data.Data[idx+1], data.Data[idx+2]

License

MIT