Develop Reference

How to draw lines unaffected by scaling and zooming

I'm trying to build an app where user can add various objects (rectangles, circles) and he can use mouse wheel to zoom-in and zoom-out.
For this zooming I set up event handler like this:
TheCanvas.on('mouse:wheel', function(options){
var p = new fabric.Point(
options.e.clientX,
options.e.clientY
);
var direction = (options.e.deltaY > 0) ? 0.9 : 1.1;
var newZoom = TheCanvas.getZoom() * direction;
// restrict too big/small zoom here:
if ((newZoom > 50) || (newZoom < 0.7)) return false;
TheCanvas.zoomToPoint( p, newZoom );
}
Everything worked fine until now. Now I want to draw a crosshair over all objects on the canvas. Something like this:
So I made my own custom object like:
CrossHairClass = fabric.util.createClass(fabric.Object, {
strokeDashArray: [1,2], // I want lines to be dashed
........
My problem is:
When user zooms with the mouse wheel, my cross-hair lines zoom their thickness too and also small dashes get bigger. But I don't want that. I want my cross-hair lines be a "hair" lines = ideally 1 pixel thick all the time regardless zoom factor of the canvas. And fine dashed line too.
Render function of my Class:
_render: function (ctx) {
// I tried it like this
var zoom = TheCanvas.getZoom();
var scale = (1/zoom) * 3.333; // with this scale it visually looked the best
// I have to scale it in X and Y while I want small dashes to stay small and also thickness of the line to stay "hair-line"
this.scaleX = this.scaleY = scale;
this.width = CROSSHAIR_SIZE / scale; // my constant from elsewhere
ctx.lineWidth = 1;
ctx.beginPath();
// this example is for horizontal line only
ctx.moveTo(-this.width / 2, 0);
ctx.lineTo(this.width / 2, 0);
this._renderStroke(ctx);
}
I tried various combinations of multiplying or dividing by scale factor or zoom factor but if I finally had lines thin, I couldn't keep their size, which must be constant (in pixels) regardless of canvas zoom. Please help.
P.S.: now I got an idea. Maybe I should create another canvas, over my current canvas and draw this crosshair on the upper canvas, which will not zoom?
EDIT 1
Based on the answer from #andreabogazzi I tried various approaches, but this finally worked out! Thanks! :)
_render: function (ctx) {
var zoom = TheCanvas.getZoom();
// ctx.save(); // this made no difference
// ctx.setTransform(1/zoom, 0, 0, 1/zoom, 0, 0); // this didn't work
this.setTransformMatrix([1/zoom, 0, 0, 1/zoom, 0, 0]);
ctx.strokStyle = 'red';
ctx.lineWidth = 1;
ctx.beginPath();
ctx.moveTo(-this.widthHalf, 0); // widthHalf computed elsewhere
ctx.lineTo(this.widthHalf, 0);
this._renderStroke(ctx); // I use this instead of ctx.stroke() while this ensures my line is still nicely dashed
// ctx.restore(); // this made no difference
}

Since you created a custom class, you have to invert the zoom of your canvas before drawing.
On the _render function of your subclass, since you should be positioned in the center of your crosshair, apply a transform matrix of scale type, with scale factor of 1/zoomLevel and everything should work.
I would say the correct way is:
_render: function (ctx) {
var zoom = TheCanvas.getZoom();
ctx.save(); // this is done anyway but if you add custom ctx transform is good practice to wrap it in a save/restore couple
ctx.transform(1/zoom, 0, 0, 1/zoom, 0, 0);
ctx.strokStyle = 'red';
ctx.lineWidth = 1;
ctx.beginPath();
ctx.moveTo(-this.widthHalf, 0); // widthHalf computed elsewhere
ctx.lineTo(this.widthHalf, 0);
this._renderStroke(ctx); // I use this instead of ctx.stroke() while this ensures my line is still nicely dashed
ctx.restore(); // this is done anyway but if you add custom ctx transform is good practice to wrap it in a save/restore couple
}
Now it happens that this object get cached from the fabricJS cache system that will probably create the cache depending on the canvas zoom too.
I have no understanding of the final use of this object, but you should include this calculation also in the cache canvas size calculation.

Slitting HTML5 canvas (video) element into pieces

I am new with canvas and I've been Googling for a couple of hours, but I am stuck.
What I would like to do is to render a video on a canvas element, divide it and animate the pieces. I am halfway there (see: http://jsbin.com/riduxadazi/edit?html,css,js,console,output ) but I have a couple of questions:
Am I doing things right, or is this extremly inefficient?
I would like to use the video fullscreen. Whatever I try, the canvas grid + video don't seem to match size.
I would like to animate the pieces of the video, but I have no clue how I should address them. Can I get some sort of array and animate the pieces one by one?
My JS looks like this. I tried to add comments to the most important parts. At least what I think were the most important parts ;)
var video = document.getElementById('video'); // Get the video
var ctx = canvas.getContext('2d'),
columns = 6,
rows = 4,
w, h, tileWidth, tileHeight;
// Start video and add it to canvas
video.addEventListener('play', function() {
var $this = this; //cache
(function loop() {
if (!$this.paused && !$this.ended) {
ctx.drawImage($this, 0, 0,window.innerWidth,window.innerHeight);
calcSize(); // Divide video
setTimeout(loop, 1000 / 30); // drawing at 30fps
}
})();
}, 0);
function calcSize() {
video.width = w = window.innerWidth;
video.height = h = window.innerHeight;
tileWidth = w / columns;
tileHeight = h / rows;
ctx.strokeStyle = '#000';
render();
}
function render() {
for(var x = 0; x < columns; x++) {
ctx.moveTo(x * tileWidth, 0);
ctx.lineTo(x * tileWidth, h);
}
for(var y = 0; y < rows; y++) {
ctx.moveTo(0, y * tileHeight);
ctx.lineTo(w, y * tileHeight);
}
ctx.stroke();
}

You would perhaps consider:
Using requestAnimationFrame to update the loop. This allows for perfect synchronization with the monitor update rate as well as being more efficient than setTimeout/setInterval You could throttle it so you only update per 1/30 frame to match video rate by using a simple boolean flag that alternates.
The video element does not need to be inserted into DOM. Also, the actual video bitmap size is read through the properties videoWidth and videoHeight, though, in the provided code you should use canvas' properties width and height as this determine the destination size. To draw proportional you can for example use this answer.
Using drawImage() using the clipping parameters would be the more efficient way to draw video onto canvas if you want to split the content.
You could split your video using a mathematical approach (see this answer) or using objects which allows you to define source regions and have individual properties on it such as position, rotation, scale and so forth. In case you would have to consider destination position to adopt to the current size of canvas.

Fill image with texture/pattern

I'm looking for a solution to change the texture/pattern for a product.
At this moment i have:
A .png picture of a couch with a transparent background
A .png picture of a texture
With the following code:
<canvas id="a" width="800" height="500">Canvas not supported on your browser</canvas>
var width = $(window).width();
var height = $(window).height();
var c = document.getElementById("a");
var ctx = c.getContext("2d");
var can2 = document.createElement('canvas');
document.body.appendChild(can2)
can2.width = c.width;
can2.height = c.height;
var ctx2 = can2.getContext("2d");
var test = new Image();
test.src = "Images/newBank.png";
test.onload = function () {
ctx2.drawImage(test, 0, 0);
};
var img = new Image();
img.src = "Images/texturetrans.png";
img.onload = function () {
ctx2.globalCompositeOperation = 'source-in';
var ptrn = ctx2.createPattern(img, 'repeat');
ctx2.fillStyle = ptrn;
ctx2.fillRect(0, 0, can2.width, can2.height);
}
`
I get this result:
As you can see, the whole object is filled with my texture. No definitions of the pillows etc. are visible anymore. Is it possible to let my texture be a sort of transparent mask?
I'm already able to change the color of the couch:
But I'd like to be able to also add a pattern to my couch!
Any help will be appreciated and I'm already very sorry for my bad English.

If you're just after an illustrative approximation you can use a combination of blending and composition modes.
First thing is to make sure your main image has transparency - this is important for composition to work (I made a rough cut-off in the following demo).
Main steps:
Draw the pattern
Draw the main image on top with blending mode multiply
Draw the main image on top with compositing mode destination-in - this will make a cut-out
If you want to reduce the size of the pattern you can either do this by using a smaller version of the image, draw to a temporary canvas at a smaller size and use that as pattern, or use the new transform methods on the pattern itself.
Demo
var img1 = new Image, img2 = new Image, cnt = 2,
canvas = document.getElementById("canvas"),
ctx = canvas.getContext("2d");
// image loading for demo (ignore)
img1.onload = img2.onload = function() {if (!--cnt) go()};
img1.src = "//i.imgur.com/8WqH9v4.png"; // sofa
img2.src = "//i.stack.imgur.com/sQlu8.png"; // pattern
// MAIN CODE ---
function go() {
// create a pattern
ctx.fillStyle = ctx.createPattern(img2, "repeat");
// fill canvas with pattern
ctx.fillRect(0, 0, canvas.width, canvas.height);
// use blending mode multiply
ctx.globalCompositeOperation = "multiply";
// draw sofa on top
ctx.drawImage(img1, 0, 0, img1.width*.5, img1.height*.5);
// change composition mode
ctx.globalCompositeOperation = "destination-in";
// draw to cut-out sofa
ctx.drawImage(img1, 0, 0, img1.width*.5, img1.height*.5);
}
<canvas id="canvas" width=600 height=400></canvas>
You can also reverse the order of which image is drawn etc., if you prefer. This is just an example of one way.
If you need accurate texture then there is no way around to either take photos or use a 3D software, or hand-drawn the textures.
NOTE: IE does not support multiply - For this you need to manually iterate through the pixels and multiply each component with each other.
You can test for support this way:
ctx.globalCompositeOperation = "multiply";
if (ctx.globalCompositeOperation === "multiply") {
// blend as above
}
else {
// iterate and blend manually
}
Blending mode luminosity is mentioned in comments and this can be used too of course. I just want to point a couple of things to consider. The first is that this is a non-separable blending mode meaning it depends on all components as it goes through the HSL color model. This makes it a bit more compute intensive.
The second is that if you end up having to do this manually (in for example IE) the code is a bit more complex to emulate, and will be noticeably slower.

Any javascript method for canvas image cover detection?

I'm developing following case.
Html canvas created by JS.
There are a star image(loaded from png with alpha=0 background) and a diamond image(also loaded from png with alpha=0 background) on canvas.
diamond image is moving toward the star image.
when diamond image is completely behind the star image, like showing star image only and diamond image is completely behind the star image, alert("Hidden");
if more than one pixel of the diamond is shown, alert should not appear.
Since the alpha value of the background of the star is 0, which means star is not a rectangle, it is difficult to detect whether the star image is fully covering the diamond image.
Is there any library or way to detect whether an image is fully covered by other?
Or, does any one know the name of this algorithm so that I can implement in JS?
Thanks for any help!

For objects with unknown shape we can check if object is behind by using pixel check.
Here is a full example on how to do this:
ONLINE DEMO HERE
(GameAlchemist provided a modified version here)
/// basic allocations
var ctx = demo.getContext('2d'),
os = document.createElement('canvas'),
octx = os.getContext('2d'),
w = os.width = demo.width,
h = os.height = demo.height,
/// the images
urlD = 'http://i.imgur.com/U72xIMZ.png',
urlS = 'http://i.imgur.com/n5rgo11.png',
imgD = new Image(),
imgS = new Image(),
cnt = 2,
/// check region (optimized)
rect = [140, 140, 180, 60];
/// load images and when ready, start show
imgD.crossOrigin = imgS.crossOrigin = 'anonymous';
imgD.onload = imgS.onload = function() {
cnt--;
if (cnt === 0) start();
}
imgD.src = urlD;
imgS.src = urlS;
The main function checks the pixels within the region defined above. To optimize we can narrow down the search area. If you need to check if image is visible on the other size the region is simply extended to check that area as well.
The function compares an off-screen canvas with just the foremost image drawn against the "live" canvas where both background and foreground are drawn.
If live canvas = off-screen canvas that means the background image is not visible.
function start() {
octx.drawImage(imgS, (w - imgS.width) * 0.5, 20);
var x = -50,
buffer1 = octx.getImageData(rect[0], rect[1], rect[2], rect[3]).data,
len = buffer1.length;
loop();
function loop() {
ctx.clearRect(0, 0, w, h);
ctx.drawImage(imgD, x, 130);
ctx.drawImage(imgS, (w - imgS.width) * 0.5, 20);
if (compare() === true) {
info.innerHTML = 'Object is behind!';
return;
}
x += 2;
if (x < w) requestAnimationFrame(loop);
}
function compare() {
var buffer2 = ctx.getImageData(rect[0], rect[1], rect[2], rect[3]).data,
i = len - 1;
while(i--) {
if (buffer1[i] !== buffer2[i]) return false
}
return true;
}
}

Find height of rendered/visible text

I know how to get this height of a font:
By placing the text in a div and getting offset height of the div.
But I would like to get this actual height (Which will depend on font family):
Is that in any way possible using web based programming?

Is there a simple solution? I think the answer is no.
If you're ok with a more involved (and processor-intensive) solution, you could try this:
Render the text to a canvas, then use canvasCtx.getImageData(..) to retrieve pixel information. Next you would do something similar to what this pseudo code describes:
first_y : null
last_y : null
for each y:
for each x:
if imageData[x][y] is black:
if first_y is null:
first_y = y
last_y = y
height = last_y - first_y
This basically looks for the top (lowest y-index) of the lettering (black pixels) and the bottom (highest y-index) then subtracts to retrieve the height.

I was writing the code while Jason answered, but I decided to post it anyway:
http://jsfiddle.net/adtn8/2/
If you follow the comments you should get the idea what's going on and why. It works pretty fast and it's not so complicated as it may sound. Checked with GIMP and it is accurate.
(code to be sure it wont be lost):
// setup variables
var c = document.createElement('canvas'),
div = document.getElementsByTagName('div')[0],
out = document.getElementsByTagName('output')[0];
// set canvas's size to be equal with div
c.width = div.offsetWidth;
c.height = div.offsetHeight;
var ctx = c.getContext('2d');
// get div's font from computed style and apply it to context
ctx.font = window.getComputedStyle(div).font;
// use color other than black because all pixels are 0 when black and transparent
ctx.fillStyle = '#bbb';
// draw the text near the bottom of the canvas
ctx.fillText(div.innerText, 0, div.offsetHeight);
// loop trough the canvas' data to find first colored pixel
var data = ctx.getImageData(0, 0, c.width, c.height).data,
minY = 0, len = data.length;
for (var i = 0; i < len; i += 4) {
// when you found it
if (data[i] != 0) {
// get pixel's y position
minY = Math.floor(i / 4 / c.width);
break;
}
}
// and print out the results
out.innerText = c.height - minY + 'px';
EDIT:
I even made jQuery plugin for this: https://github.com/maciek134/jquery-textHeight
Enjoy.