사진 (Sajin)

Sajin is an easy to use library (specially designed for Selenium) aimed at comparing images and their similarity (as a double data type; representing percentage).

Add Sajin to your project

Step 1: Add the JitPack repository to your build file

<repositories>
    <repository>
        <id>jitpack.io</id>
        <url>https://jitpack.io</url>
    </repository>
</repositories>

Step 2: Add the dependency

<dependency>
    <groupId>com.github.memburg</groupId>
    <artifactId>sajin</artifactId>
    <version>1.0.5</version>
</dependency>

How to use

// Take screenshot #1
String expected = ((TakesScreenshot) driver).getScreenshotAs(OutputType.BASE64);

// Some other logic in between (e.g. clicks and/or scrolls)

// Take screenshot #2
String actual = ((TakesScreenshot) driver).getScreenshotAs(OutputType.BASE64);

// Make the comparison!
double similarity = Sajin.hardCompare(expected, actual);

// Print the results
System.out.println("Similarity is " + similarity + "%");
// Expected output: Similarity is 96.60064%

Methods

Sajin is a small library, it only has FOUR methods:

• softCompare(expected: String, actual: String, standardDeviation: Int): Float
• midCompare(expected: String, actual: String, standardDeviation: Int): Float
• hardCompare(expected: String, actual: String): Float
• detectMovement(i1: String, 12: String, r: Int, g: Int, b: Int): String

The first three methods compare images, simple as that, without going too deep into the algorithms, softCompare and midCompare take a standard deviation as a parameter to avoid evaluation as strict as hardCompare.

hardCompare uses a pixel-by-pixel comparison.

detectMovement uses a pixel-by-pixel matching algorithm which returns a BASE64 image with the pixels that were detected as different marked with the color given by the user in the method call.

Important

In version 1.0.2 it is necessary to use Sajin like in the following example:

// Preferably use version 1.0.5
double similarity = Sajin.INSTANCE.hardCompare(expected, actual);

Contributions

This is an open source project, feel free to contribute to it!