imagecolorclosest

(PHP 4, PHP 5, PHP 7, PHP 8)

imagecolorclosest — Get the index of the closest color to the specified color

Description

imagecolorclosest(
    GdImage $image,
    int $red,
    int $green,
    int $blue
): int

Returns the index of the color in the palette of the image which is "closest" to the specified RGB value.

The "distance" between the desired color and each color in the palette is calculated as if the RGB values represented points in three-dimensional space.

If you created the image from a file, only colors used in the image are resolved. Colors present only in the palette are not resolved.

Parameters

image: A GdImage object, returned by one of the image creation functions, such as imagecreatetruecolor().
red: Value of red component.
green: Value of green component.
blue: Value of blue component.

The colors parameters are integers between 0 and 255 or hexadecimals between 0x00 and 0xFF.

Return Values

Returns the index of the closest color, in the palette of the image, to the specified one

Changelog

Version	Description
8.0.0	`image` expects a GdImage instance now; previously, a valid `gd` resource was expected.

Examples

Example #1 Search for a set of colors in an image

<?php
// Start with an image and convert it to a palette-based image
$im = imagecreatefrompng('figures/imagecolorclosest.png');
imagetruecolortopalette($im, false, 255);

// Search colors (RGB)
$colors = array(
    array(254, 145, 154),
    array(153, 145, 188),
    array(153, 90, 145),
    array(255, 137, 92)
);

// Loop through each search and find the closest color in the palette.
// Return the search number, the search RGB and the converted RGB match
foreach($colors as $id => $rgb)
{
$result = imagecolorclosest($im, $rgb[0], $rgb[1], $rgb[2]);
$result = imagecolorsforindex($im, $result);
$result = "({$result['red']}, {$result['green']}, {$result['blue']})";

    echo "#$id: Search ($rgb[0], $rgb[1], $rgb[2]); Closest match: $result.\n";
}

imagedestroy($im);
?>

The above example will output something similar to:

#0: Search (254, 145, 154); Closest match: (252, 150, 148).
#1: Search (153, 145, 188); Closest match: (148, 150, 196).
#2: Search (153, 90, 145); Closest match: (148, 90, 156).
#3: Search (255, 137, 92); Closest match: (252, 150, 92).

Improve This Page

Learn How To Improve This Page • Submit a Pull Request • Report a Bug

＋add a note

User Contributed Notes 4 notes

down

info at codeworx dot ch ¶

12 years ago

Here is a function that compares two HEX colors for similarity. This can be useful if you want to detect colors that are not different enough to see for the naked eye. It returns a bool: TRUE if the colors are similar to each other (within tolerance) or FALSE if they are different enough.
I tested a few colors and came up with a tolerance of 35 (rgb value - should be between 0 and 255) but you can change that tolerance by passing a third parameter to the function.

<?php
function compareColors ($col1, $col2, $tolerance=35) {
$col1Rgb = array(
"r" => hexdec(substr($col1, 0, 2)),
"g" => hexdec(substr($col1, 2, 2)),
"b" => hexdec(substr($col1, 4, 2))
    );
$col2Rgb = array(
"r" => hexdec(substr($col2, 0, 2)),
"g" => hexdec(substr($col2, 2, 2)),
"b" => hexdec(substr($col2, 4, 2))
    );
    return ($col1Rgb['r'] >= $col2Rgb['r'] - $tolerance && $col1Rgb['r'] <= $col2Rgb['r'] + $tolerance) && ($col1Rgb['g'] >= $col2Rgb['g'] - $tolerance && $col1Rgb['g'] <= $col2Rgb['g'] + $tolerance) && ($col1Rgb['b'] >= $col2Rgb['b'] - $tolerance && $col1Rgb['b'] <= $col2Rgb['b'] + $tolerance);
}
?>

down

Hayley Watson ¶

7 years ago

RGB space isn't the best choice for measuring the distance between two colours, since it ignores, for example, the fact that we count both dark green and light green as "green" (the RGB distance between #000000 and #7f7f7f is the same as the distance between #000000 and #5443c0 - a slightly darkened SlateBlue).

A better choice of colour space that conforms better to how colours are perceived is the so-called Lab space, which measures colours according to how light/dark, red/green, and yellow/blue they are. (There are still better models, but they come at the expense of increased computation.)

<?php

function warp1($c)
{
    if($c > 10.3148)
    {
        return pow((561 + 40*$c)/10761, 2.4);
    }
    else
    {
        return $c / 3294.6;
    }
}
function warp2($c)
{
    if($c > 0.008856)
    {
        return pow($c, 1/3);
    }
    else
    {
        return 7.787 * $c + 4/29;
    }
}
function rgb2lab($rgb)
{
    [$red, $green, $blue] = array_map('warp1', $rgb);

$x = warp2($red * 0.4339 + $green * 0.3762 + $blue * 0.1899);
$y = warp2($red * 0.2126 + $green * 0.7152 + $blue * 0.0722);
$z = warp2($red * 0.0178 + $green * 0.1098 + $blue * 0.8730);

$l = 116*$y - 16;
$a = 500 * ($x - $y);
$b = 200 * ($y - $z);

    return array_map('intval', [$l, $a, $b]);
}

function generate_palette_from_image($image)
{
$pal = [];
$width = imagesx($image);
$height = imagesy($image);
    for($x = 0; $x < $width; ++$x)
    {
        for($y = 0; $y < $height; ++$y)
        {
$pal[] = imagecolorat($image, $x, $y);
        }
    }
    return array_map(function($col)use($image)
    {
$rgba = imagecolorsforindex($image, $col);
        return [$rgba['red'], $rgba['green'], $rgba['blue']];
    },    array_unique($pal));
}

function closest_rgb_in_palette($rgb, $palette)
{
// Quick return when the exact
    // colour is in the palette.
if(($idx = array_search($rgb, $palette)) !== false)
    {
        return $idx;
    }
    [$tl, $ta, $tb] = rgb2lab($rgb);
$dists = array_map(function($plab)use($tl, $ta, $tb)
    {
        [$pl, $pa, $pb] = $plab;
$dl = $pl - $tl;
$da = $pa - $ta;
$db = $pa - $tb;
        return $dl * $dl + $da * $da + $db * $db;
    }, array_map('rgb2lab', $palette));
    return array_search(min($dists), $dists);
}

function closest_rgb_in_image($rgb, $image)
{
$palette = generate_palette_from_image($image);
    return $palette[closest_rgb_in_palette($rgb, $palette)];
}

$lena = imagecreatefrompng('lena.png');
$red = closest_rgb_in_image([255,0,0],$lena);
echo join(' ', $red);  // 228 71 82

?>

If you're going to be matching a lot of colours to a palette, you may want to precompute and reuse the Lab palette, instead of generating it fresh each time as done here.

down

-1

MagicalTux at FF dot st ¶

19 years ago

A way to get each time an answer :

<?php
function imagegetcolor($im, $r, $g, $b) {
$c=imagecolorexact($im, $r, $g, $b);
        if ($c!=-1) return $c;
$c=imagecolorallocate($im, $r, $g, $b);
        if ($c!=-1) return $c;
        return imagecolorclosest($im, $r, $g, $b);
}
?>

If the *exact* color is found in the image, it will be returned. If we don't have the exact color, we try to allocate it. If we can't allocate it, we return the closest color in the image.

down

-1

Vikrant Korde <vakorde at hotmail dot com> ¶

20 years ago

This functuion is useful when you are dealing with previously present images like .png, .jpg, etc. You can use this function for writing a text on the image.

For me the function imagecolorallocate() was returning the -1 value. after lot of RnD and site search i found a function use of imagecolorclosest(). This solved my problem of writing the text on the image with customised color. 

Actually previously it was writing on the image but the color was not distinct. It was using the same color as of that background image. 

The following code segment was fine with me.

header ("Content-type: image/jpeg"); 
$im = imagecreatefromjpeg("BlankButton.jpg");
$white = imageColorClosest($im, 255,255,255);
// this is for TTF fonts
imagettftext ($im, 20, 0, 16, 30, $white, "/usr/X11R6/lib/X11/fonts/TTF/luximb.ttf", "Vikrant"); 

//this is for notmal font
imagestring($im, 4, 0,0,"Korde", $white);
imagejpeg($im,"",150); 
imagedestroy ($im);

＋add a note

imagecolorclosest

Description

Parameters

Return Values

Changelog

Examples

See Also

Improve This Page

User Contributed Notes 4 notes