Home

Awesome

Imaging

PkgGoDev Build Status Coverage Status Go Report Card

Package imaging provides basic image processing functions (resize, rotate, crop, brightness/contrast adjustments, etc.).

All the image processing functions provided by the package accept any image type that implements image.Image interface as an input, and return a new image of *image.NRGBA type (32bit RGBA colors, non-premultiplied alpha).

Installation

go get -u github.com/disintegration/imaging

Documentation

https://pkg.go.dev/github.com/disintegration/imaging

Usage examples

A few usage examples can be found below. See the documentation for the full list of supported functions.

Image resizing

// Resize srcImage to size = 128x128px using the Lanczos filter.
dstImage128 := imaging.Resize(srcImage, 128, 128, imaging.Lanczos)

// Resize srcImage to width = 800px preserving the aspect ratio.
dstImage800 := imaging.Resize(srcImage, 800, 0, imaging.Lanczos)

// Scale down srcImage to fit the 800x600px bounding box.
dstImageFit := imaging.Fit(srcImage, 800, 600, imaging.Lanczos)

// Resize and crop the srcImage to fill the 100x100px area.
dstImageFill := imaging.Fill(srcImage, 100, 100, imaging.Center, imaging.Lanczos)

Imaging supports image resizing using various resampling filters. The most notable ones:

The full list of supported filters: NearestNeighbor, Box, Linear, Hermite, MitchellNetravali, CatmullRom, BSpline, Gaussian, Lanczos, Hann, Hamming, Blackman, Bartlett, Welch, Cosine. Custom filters can be created using ResampleFilter struct.

Resampling filters comparison

Original image:

srcImage

The same image resized from 600x400px to 150x100px using different resampling filters. From faster (lower quality) to slower (higher quality):

FilterResize result
imaging.NearestNeighbordstImage
imaging.LineardstImage
imaging.CatmullRomdstImage
imaging.LanczosdstImage

Gaussian Blur

dstImage := imaging.Blur(srcImage, 0.5)

Sigma parameter allows to control the strength of the blurring effect.

Original imageSigma = 0.5Sigma = 1.5
srcImagedstImagedstImage

Sharpening

dstImage := imaging.Sharpen(srcImage, 0.5)

Sharpen uses gaussian function internally. Sigma parameter allows to control the strength of the sharpening effect.

Original imageSigma = 0.5Sigma = 1.5
srcImagedstImagedstImage

Gamma correction

dstImage := imaging.AdjustGamma(srcImage, 0.75)
Original imageGamma = 0.75Gamma = 1.25
srcImagedstImagedstImage

Contrast adjustment

dstImage := imaging.AdjustContrast(srcImage, 20)
Original imageContrast = 15Contrast = -15
srcImagedstImagedstImage

Brightness adjustment

dstImage := imaging.AdjustBrightness(srcImage, 20)
Original imageBrightness = 10Brightness = -10
srcImagedstImagedstImage

Saturation adjustment

dstImage := imaging.AdjustSaturation(srcImage, 20)
Original imageSaturation = 30Saturation = -30
srcImagedstImagedstImage

Hue adjustment

dstImage := imaging.AdjustHue(srcImage, 20)
Original imageHue = 60Hue = -60
srcImagedstImagedstImage

FAQ

Incorrect image orientation after processing (e.g. an image appears rotated after resizing)

Most probably, the given image contains the EXIF orientation tag. The standard image/* packages do not support loading and saving this kind of information. To fix the issue, try opening images with the AutoOrientation decode option. If this option is set to true, the image orientation is changed after decoding, according to the orientation tag (if present). Here's the example:

img, err := imaging.Open("test.jpg", imaging.AutoOrientation(true))

What's the difference between imaging and gift packages?

imaging is designed to be a lightweight and simple image manipulation package. It provides basic image processing functions and a few helper functions such as Open and Save. It consistently returns *image.NRGBA image type (8 bits per channel, RGBA).

gift supports more advanced image processing, for example, sRGB/Linear color space conversions. It also supports different output image types (e.g. 16 bits per channel) and provides easy-to-use API for chaining multiple processing steps together.

Example code

package main

import (
	"image"
	"image/color"
	"log"

	"github.com/disintegration/imaging"
)

func main() {
	// Open a test image.
	src, err := imaging.Open("testdata/flowers.png")
	if err != nil {
		log.Fatalf("failed to open image: %v", err)
	}

	// Crop the original image to 300x300px size using the center anchor.
	src = imaging.CropAnchor(src, 300, 300, imaging.Center)

	// Resize the cropped image to width = 200px preserving the aspect ratio.
	src = imaging.Resize(src, 200, 0, imaging.Lanczos)

	// Create a blurred version of the image.
	img1 := imaging.Blur(src, 5)

	// Create a grayscale version of the image with higher contrast and sharpness.
	img2 := imaging.Grayscale(src)
	img2 = imaging.AdjustContrast(img2, 20)
	img2 = imaging.Sharpen(img2, 2)

	// Create an inverted version of the image.
	img3 := imaging.Invert(src)

	// Create an embossed version of the image using a convolution filter.
	img4 := imaging.Convolve3x3(
		src,
		[9]float64{
			-1, -1, 0,
			-1, 1, 1,
			0, 1, 1,
		},
		nil,
	)

	// Create a new image and paste the four produced images into it.
	dst := imaging.New(400, 400, color.NRGBA{0, 0, 0, 0})
	dst = imaging.Paste(dst, img1, image.Pt(0, 0))
	dst = imaging.Paste(dst, img2, image.Pt(0, 200))
	dst = imaging.Paste(dst, img3, image.Pt(200, 0))
	dst = imaging.Paste(dst, img4, image.Pt(200, 200))

	// Save the resulting image as JPEG.
	err = imaging.Save(dst, "testdata/out_example.jpg")
	if err != nil {
		log.Fatalf("failed to save image: %v", err)
	}
}

Output:

dstImage