Has anyone implemented a flood fill algorithm in javascript for use with HTML Canvas?
My requirements are simple: flood with a single color starting from a single po
I would not treat the canvas as a bitmap image.
Instead I would keep a collection of painting-objects and modify that collection. Then for example you can fill a path or shape or add a new shape that has the boundaries of the objects you are trying to fill.
I can't see how "normal" floodFill makes sense in vector drawing..
Here's an implementation that I've been working on. It can get really slow if the replacement color is too close to the original color. It's quite a bit faster in Chrome than Firefox (I haven't tested it in any other browsers).
I also haven't done exhaustive testing yet, so there may be edge cases where it doesn't work.
function getPixel(pixelData, x, y) {
if (x < 0 || y < 0 || x >= pixelData.width || y >= pixelData.height) {
return NaN;
}
var pixels = pixelData.data;
var i = (y * pixelData.width + x) * 4;
return ((pixels[i + 0] & 0xFF) << 24) |
((pixels[i + 1] & 0xFF) << 16) |
((pixels[i + 2] & 0xFF) << 8) |
((pixels[i + 3] & 0xFF) << 0);
}
function setPixel(pixelData, x, y, color) {
var i = (y * pixelData.width + x) * 4;
var pixels = pixelData.data;
pixels[i + 0] = (color >>> 24) & 0xFF;
pixels[i + 1] = (color >>> 16) & 0xFF;
pixels[i + 2] = (color >>> 8) & 0xFF;
pixels[i + 3] = (color >>> 0) & 0xFF;
}
function diff(c1, c2) {
if (isNaN(c1) || isNaN(c2)) {
return Infinity;
}
var dr = ((c1 >>> 24) & 0xFF) - ((c2 >>> 24) & 0xFF);
var dg = ((c1 >>> 16) & 0xFF) - ((c2 >>> 16) & 0xFF);
var db = ((c1 >>> 8) & 0xFF) - ((c2 >>> 8) & 0xFF);
var da = ((c1 >>> 0) & 0xFF) - ((c2 >>> 0) & 0xFF);
return dr*dr + dg*dg + db*db + da*da;
}
function floodFill(canvas, x, y, replacementColor, delta) {
var current, w, e, stack, color, cx, cy;
var context = canvas.getContext("2d");
var pixelData = context.getImageData(0, 0, canvas.width, canvas.height);
var done = [];
for (var i = 0; i < canvas.width; i++) {
done[i] = [];
}
var targetColor = getPixel(pixelData, x, y);
delta *= delta;
stack = [ [x, y] ];
done[x][y] = true;
while ((current = stack.pop())) {
cx = current[0];
cy = current[1];
if (diff(getPixel(pixelData, cx, cy), targetColor) <= delta) {
setPixel(pixelData, cx, cy, replacementColor);
w = e = cx;
while (w > 0 && diff(getPixel(pixelData, w - 1, cy), targetColor) <= delta) {
--w;
if (done[w][cy]) break;
setPixel(pixelData, w, cy, replacementColor);
}
while (e < pixelData.width - 1 && diff(getPixel(pixelData, e + 1, cy), targetColor) <= delta) {
++e;
if (done[e][cy]) break;
setPixel(pixelData, e, cy, replacementColor);
}
for (cx = w; cx <= e; cx++) {
if (cy > 0) {
color = getPixel(pixelData, cx, cy - 1);
if (diff(color, targetColor) <= delta) {
if (!done[cx][cy - 1]) {
stack.push([cx, cy - 1]);
done[cx][cy - 1] = true;
}
}
}
if (cy < canvas.height - 1) {
color = getPixel(pixelData, cx, cy + 1);
if (diff(color, targetColor) <= delta) {
if (!done[cx][cy + 1]) {
stack.push([cx, cy + 1]);
done[cx][cy + 1] = true;
}
}
}
}
}
}
context.putImageData(pixelData, 0, 0, 0, 0, canvas.width, canvas.height);
}
To create a flood fill you need to be able to look at the pixels that are there already and check they aren't the color you started with so something like this.
const ctx = document.querySelector("canvas").getContext("2d");
ctx.beginPath();
ctx.moveTo(20, 20);
ctx.lineTo(250, 70);
ctx.lineTo(270, 120);
ctx.lineTo(170, 140);
ctx.lineTo(190, 80);
ctx.lineTo(100, 60);
ctx.lineTo(50, 130);
ctx.lineTo(20, 20);
ctx.stroke();
floodFill(ctx, 40, 50, [255, 0, 0, 255]);
function getPixel(imageData, x, y) {
if (x < 0 || y < 0 || x >= imageData.width || y >= imageData.height) {
return [-1, -1, -1, -1]; // impossible color
} else {
const offset = (y * imageData.width + x) * 4;
return imageData.data.slice(offset, offset + 4);
}
}
function setPixel(imageData, x, y, color) {
const offset = (y * imageData.width + x) * 4;
imageData.data[offset + 0] = color[0];
imageData.data[offset + 1] = color[1];
imageData.data[offset + 2] = color[2];
imageData.data[offset + 3] = color[0];
}
function colorsMatch(a, b) {
return a[0] === b[0] && a[1] === b[1] && a[2] === b[2] && a[3] === b[3];
}
function floodFill(ctx, x, y, fillColor) {
// read the pixels in the canvas
const imageData = ctx.getImageData(0, 0, ctx.canvas.width, ctx.canvas.height);
// get the color we're filling
const targetColor = getPixel(imageData, x, y);
// check we are actually filling a different color
if (!colorsMatch(targetColor, fillColor)) {
fillPixel(imageData, x, y, targetColor, fillColor);
// put the data back
ctx.putImageData(imageData, 0, 0);
}
}
function fillPixel(imageData, x, y, targetColor, fillColor) {
const currentColor = getPixel(imageData, x, y);
if (colorsMatch(currentColor, targetColor)) {
setPixel(imageData, x, y, fillColor);
fillPixel(imageData, x + 1, y, targetColor, fillColor);
fillPixel(imageData, x - 1, y, targetColor, fillColor);
fillPixel(imageData, x, y + 1, targetColor, fillColor);
fillPixel(imageData, x, y - 1, targetColor, fillColor);
}
}
<canvas></canvas>
There's at least 2 problems with this code though.
It's deeply recursive.
So you might run out of stack space
It's slow.
No idea if it's too slow but JavaScript in the browser is mostly single threaded so while this code is running the browser is frozen. For a large canvas that frozen time might make the page really slow and if it's frozen too long the browser will ask if the user wants to kill the page.
The solution to running out of stack space is to implement our own stack. For example instead of recursively calling fillPixel
we could keep an array of positions we want to look at. We'd add the 4 positions to that array and then pop things off the array until it's empty
const ctx = document.querySelector("canvas").getContext("2d");
ctx.beginPath();
ctx.moveTo(20, 20);
ctx.lineTo(250, 70);
ctx.lineTo(270, 120);
ctx.lineTo(170, 140);
ctx.lineTo(190, 80);
ctx.lineTo(100, 60);
ctx.lineTo(50, 130);
ctx.lineTo(20, 20);
ctx.stroke();
floodFill(ctx, 40, 50, [255, 0, 0, 255]);
function getPixel(imageData, x, y) {
if (x < 0 || y < 0 || x >= imageData.width || y >= imageData.height) {
return [-1, -1, -1, -1]; // impossible color
} else {
const offset = (y * imageData.width + x) * 4;
return imageData.data.slice(offset, offset + 4);
}
}
function setPixel(imageData, x, y, color) {
const offset = (y * imageData.width + x) * 4;
imageData.data[offset + 0] = color[0];
imageData.data[offset + 1] = color[1];
imageData.data[offset + 2] = color[2];
imageData.data[offset + 3] = color[0];
}
function colorsMatch(a, b) {
return a[0] === b[0] && a[1] === b[1] && a[2] === b[2] && a[3] === b[3];
}
function floodFill(ctx, x, y, fillColor) {
// read the pixels in the canvas
const imageData = ctx.getImageData(0, 0, ctx.canvas.width, ctx.canvas.height);
// get the color we're filling
const targetColor = getPixel(imageData, x, y);
// check we are actually filling a different color
if (!colorsMatch(targetColor, fillColor)) {
const pixelsToCheck = [x, y];
while (pixelsToCheck.length > 0) {
const y = pixelsToCheck.pop();
const x = pixelsToCheck.pop();
const currentColor = getPixel(imageData, x, y);
if (colorsMatch(currentColor, targetColor)) {
setPixel(imageData, x, y, fillColor);
pixelsToCheck.push(x + 1, y);
pixelsToCheck.push(x - 1, y);
pixelsToCheck.push(x, y + 1);
pixelsToCheck.push(x, y - 1);
}
}
// put the data back
ctx.putImageData(imageData, 0, 0);
}
}
<canvas></canvas>
The solution to it being too slow is either to make it run a little at a time OR to move it to a worker. I think that's a little too much to show in the same answer though here's an example. I tested the code above on a 4096x4096 canvas and it took 16 seconds to fill a blank canvas on my machine so yes it's arguably too slow but putting it in a worker brings new problems which is that the result will be asynchronous so even though the browser wouldn't freeze you'd probably want to prevent the user from doing something until it finishes.
Another issue is you'll see the lines are antialiased and so filling with a solid color fills close the the line but not all the way up to it. To fix that you can change colorsMatch
to check for close enough but then you have a new problem that if targetColor
and fillColor
are also close enough it will keep trying to fill itself. You could solve that by making another array, one byte or one bit per pixel to track places you've ready checked.
const ctx = document.querySelector("canvas").getContext("2d");
ctx.beginPath();
ctx.moveTo(20, 20);
ctx.lineTo(250, 70);
ctx.lineTo(270, 120);
ctx.lineTo(170, 140);
ctx.lineTo(190, 80);
ctx.lineTo(100, 60);
ctx.lineTo(50, 130);
ctx.lineTo(20, 20);
ctx.stroke();
floodFill(ctx, 40, 50, [255, 0, 0, 255], 128);
function getPixel(imageData, x, y) {
if (x < 0 || y < 0 || x >= imageData.width || y >= imageData.height) {
return [-1, -1, -1, -1]; // impossible color
} else {
const offset = (y * imageData.width + x) * 4;
return imageData.data.slice(offset, offset + 4);
}
}
function setPixel(imageData, x, y, color) {
const offset = (y * imageData.width + x) * 4;
imageData.data[offset + 0] = color[0];
imageData.data[offset + 1] = color[1];
imageData.data[offset + 2] = color[2];
imageData.data[offset + 3] = color[0];
}
function colorsMatch(a, b, rangeSq) {
const dr = a[0] - b[0];
const dg = a[1] - b[1];
const db = a[2] - b[2];
const da = a[3] - b[3];
return dr * dr + dg * dg + db * db + da * da < rangeSq;
}
function floodFill(ctx, x, y, fillColor, range = 1) {
// read the pixels in the canvas
const imageData = ctx.getImageData(0, 0, ctx.canvas.width, ctx.canvas.height);
// flags for if we visited a pixel already
const visited = new Uint8Array(imageData.width, imageData.height);
// get the color we're filling
const targetColor = getPixel(imageData, x, y);
// check we are actually filling a different color
if (!colorsMatch(targetColor, fillColor)) {
const rangeSq = range * range;
const pixelsToCheck = [x, y];
while (pixelsToCheck.length > 0) {
const y = pixelsToCheck.pop();
const x = pixelsToCheck.pop();
const currentColor = getPixel(imageData, x, y);
if (!visited[y * imageData.width + x] &&
colorsMatch(currentColor, targetColor, rangeSq)) {
setPixel(imageData, x, y, fillColor);
visited[y * imageData.width + x] = 1; // mark we were here already
pixelsToCheck.push(x + 1, y);
pixelsToCheck.push(x - 1, y);
pixelsToCheck.push(x, y + 1);
pixelsToCheck.push(x, y - 1);
}
}
// put the data back
ctx.putImageData(imageData, 0, 0);
}
}
<canvas></canvas>
Note that this version of colorsMatch
is using is kind of naieve. It might be better to convert to HSV or something or maybe you want to weight by alpha. I don't know what a good metric is for matching colors.
Another way to speed things up is of course to just optimize the code. Kaiido pointed out an obvious speedup which is to use a Uint32Array
view on the pixels. That way looking up a pixel and setting a pixel there's just one 32bit value to read or write. Just that change makes it about 4x faster. It still takes 4 seconds to fill a 4096x4096 canvas though. There might be other optimization like instead of calling getPixels
make that inline but don't push a new pixel on our list of pixels to check if they are out of range. It might be 10% speed up (no idea) but won't make it fast enough to be an interactive speed.
There are other speedups like checking across a row at a time since rows are cache friendly and you can compute the offset to a row once and use that while checking the entire row whereas now for every pixel we have to compute the offset multiple times.
Those will complicate the algorithm so they are best left for you to figure out.
Let me also add given the answer about freezes the browser will the fill is happen and that on a larger canvas that freeze could be too long, you can easily make the algorithm span over time using ES6 async/await. You need to choose how much work to give each segment of time. Choose too small and it will take a long time to fill. Choose too large and you'll get jank as the browser freezes.
Here's an example. Set ticksPerUpdate
to speed up or slow down the fill rate
const ctx = document.querySelector("canvas").getContext("2d");
ctx.beginPath();
ctx.moveTo(20, 20);
ctx.lineTo(250, 70);
ctx.lineTo(270, 120);
ctx.lineTo(170, 140);
ctx.lineTo(100, 145);
ctx.lineTo(110, 105);
ctx.lineTo(130, 125);
ctx.lineTo(190, 80);
ctx.lineTo(100, 60);
ctx.lineTo(50, 130);
ctx.lineTo(20, 20);
ctx.stroke();
floodFill(ctx, 40, 50, 0xFF0000FF);
function getPixel(pixelData, x, y) {
if (x < 0 || y < 0 || x >= pixelData.width || y >= pixelData.height) {
return -1; // impossible color
} else {
return pixelData.data[y * pixelData.width + x];
}
}
async function floodFill(ctx, x, y, fillColor) {
// read the pixels in the canvas
const imageData = ctx.getImageData(0, 0, ctx.canvas.width, ctx.canvas.height);
// make a Uint32Array view on the pixels so we can manipulate pixels
// one 32bit value at a time instead of as 4 bytes per pixel
const pixelData = {
width: imageData.width,
height: imageData.height,
data: new Uint32Array(imageData.data.buffer),
};
// get the color we're filling
const targetColor = getPixel(pixelData, x, y);
// check we are actually filling a different color
if (targetColor !== fillColor) {
const ticksPerUpdate = 50;
let tickCount = 0;
const pixelsToCheck = [x, y];
while (pixelsToCheck.length > 0) {
const y = pixelsToCheck.pop();
const x = pixelsToCheck.pop();
const currentColor = getPixel(pixelData, x, y);
if (currentColor === targetColor) {
pixelData.data[y * pixelData.width + x] = fillColor;
// put the data back
ctx.putImageData(imageData, 0, 0);
++tickCount;
if (tickCount % ticksPerUpdate === 0) {
await wait();
}
pixelsToCheck.push(x + 1, y);
pixelsToCheck.push(x - 1, y);
pixelsToCheck.push(x, y + 1);
pixelsToCheck.push(x, y - 1);
}
}
}
}
function wait(delay = 0) {
return new Promise((resolve) => {
setTimeout(resolve, delay);
});
}
<canvas></canvas>