# Color picker from image

Page: https://toolboxhere.com/color-picker-from-image/

Drop a photo or a screenshot and tap any pixel to get its colour. A loupe magnifies the pixels under your finger, and the colour you pick appears as hex, RGB, HSL and every other format, with its contrast and the nearest colour name.

## How to use Color Picker From Image

1. Drop a photo, a screenshot or a logo, paste one, or take a photo with your camera. JPG, PNG, WebP and iPhone HEIC photos all open.
2. Tap the colour you want, or drag across the photo: the loupe shows the pixels under your finger, magnified, with the one you will get outlined.
3. Let go to take that pixel. Its hex code and rgb() value appear beside the photo and in the converter below.
4. Copy the format you need, check the contrast on white and black, or tap a tint or shade to build a palette.

## Why use Color Picker From Image

- **A loupe for the exact pixel**: Nine by nine pixels magnified under your finger or the pointer, so you get the pixel you mean and not the one beside it.
- **Every format**: HEX, RGB, HSL, HSB, HWB, CMYK, Swift and Android, each with Copy, plus the nearest CSS colour name.
- **iPhone photos too**: HEIC photos open in every browser; where a browser cannot read them itself, they are converted on your device first.
- **Pick from the screen**: In Chrome and Edge on a computer, the eyedropper also picks any colour on your screen, outside this page too.

## About this tool

A colour picker for images reads the colour of one pixel of a picture: a brand colour from a logo, a paint shade from a photo of a wall, the exact blue of a button in a screenshot. This one works online in your browser, and the colour comes from the photo’s own pixels, read on your device. The loupe magnifies a 9 by 9 square around your finger and outlines the centre pixel, the one you get when you let go.

Photos are rarely flat colour. A wall in a photo is a spread of slightly different pixels from light, shadow and camera noise, so two taps a few pixels apart can give different codes; pick from the area that looks most typical, or compare a few. Screenshots and logos are flat, so a tap there gives the exact colour the designer used. Once picked, the colour goes into the converter, where every format, the contrast on white and black, and the tints and shades are ready to copy.

## Frequently asked questions

### Do I get the exact pixel or an average?

The exact pixel under the centre of the loupe, not an average of its neighbours. Very large photos, over 16 megapixels on a phone or 60 on a computer, are first scaled to what the browser can draw, and then one pixel stands for a few of the original.

### Why does the colour differ from what I see on my screen?

Your screen shows a photo through its own brightness, night mode and colour settings, and phones often boost colours. The code you get is the value stored for that pixel, in the sRGB colours websites and design apps use. A screenshot of a photo can differ too, because it records the screen rather than the file.

### Can I pick a colour from an iPhone photo (HEIC)?

Yes. Where your browser cannot open HEIC itself, as in Chrome, Edge and Firefox on Windows and Android, the photo is converted on your device first, which takes a moment for a large photo.

### Can it tell me the name of the colour?

Yes. The converter shows the nearest of the 148 CSS colour names, such as teal or dodgerblue, or says the colour is close to one.

### Can I get a colour from a website or another app?

Take a screenshot, drop it here and tap the colour. In Chrome and Edge on a computer, Pick from the screen takes a colour straight from anywhere on your screen.

### Can I pick a colour from something in front of me, like a wall or a fabric?

Yes. Tap "Take a photo" on a phone, or "Use the webcam" on a computer, and photograph it; the photo opens here and you tap the colour. The camera and the light change what it records, so treat the code as close to the paint or fabric rather than exact. Daylight without a flash comes nearest.

Built and maintained by Saad Naeem, AI engineer, Doha. Updated 27 October 2026.
