imghash

Perceptual image hashing for Node.js

Installation

npm install imghash

Basic usage

import imghash from "imghash";

const hash1 = await imghash.hash("./path/to/file");
console.log(hash1); // "f884c4d8d1193c07"

// Use a 4-bit hash grid and return the result as binary
const hash2 = await imghash.hash("./path/to/file", 4, "binary");
console.log(hash2); // "1000100010000010"

Finding similar images

To measure similarity between images you can use Hamming distance or Levenshtein Distance.

The following example uses the latter one:

import imghash from "imghash";
import leven from "leven";

const hash1 = await imghash.hash("./img1");
const hash2 = await imghash.hash("./img2");

const distance = leven(hash1, hash2);
console.log(`Distance between images is: ${distance}`);
if (distance <= 12) {
  console.log("Images are similar");
} else {
  console.log("Images are NOT similar");
}

The threshold of 12 is a starting point, not a universal rule. Tune it for your image set and the hash size you use.

API

hash(filepath, bits?, format?)

Calculates a perceptual hash for an image.

const hash = await imghash.hash("./image.png", 8, "hex");

Parameter	Type	Default	Description
filepath	string | Buffer	—	Image path or image data.
bits	number	8	Hash grid size. Must be divisible by 4.
format	"hex" | "binary"	"hex"	Returned hash encoding.

Returns: Promise<string>.

See the full API Reference.

Read more

• Command-Line Interface
• Block mean value hashing method