CODEWITHSUNDEEP
javascriptimage-processinghtml5-canvas
mheavers
mheavers

Reputation: 30158

Understanding how canvas converts an image to black and white

I found this script for converting an image to black and white, which works great, but I was hoping to understand the code a little bit better. I put my questions in the code, in the form of comments.

Can anyone explain in a little more detail what is happening here:

function grayscale(src){ //Creates a canvas element with a grayscale version of the color image
    var canvas = document.createElement('canvas');
    var ctx = canvas.getContext('2d');
    var imgObj = new Image();
    imgObj.src = src;
    canvas.width = imgObj.width;
    canvas.height = imgObj.height; 
    ctx.drawImage(imgObj, 0, 0); //Are these CTX functions documented somewhere where I can see what parameters they require / what those parameters mean?
    var imgPixels = ctx.getImageData(0, 0, canvas.width, canvas.height);
    for(var y = 0; y < imgPixels.height; y++){
        for(var x = 0; x < imgPixels.width; x++){
            var i = (y * 4) * imgPixels.width + x * 4; //Why is this multiplied by 4?
            var avg = (imgPixels.data[i] + imgPixels.data[i + 1] + imgPixels.data[i + 2]) / 3; //Is this getting the average of the values of each channel R G and B, and converting them to BW(?)
            imgPixels.data[i] = avg; 
            imgPixels.data[i + 1] = avg; 
            imgPixels.data[i + 2] = avg;
        }
    }
    ctx.putImageData(imgPixels, 0, 0, 0, 0, imgPixels.width, imgPixels.height); 
    return canvas.toDataURL();
}

Upvotes: 3

Views: 4788

Answers (2)

pimvdb
pimvdb

Reputation: 154838

  1. The canvas functions are, like most functions, described in an official specification. Also, MDC is helpful for more "informal" articles. E.g. the drawImage function on MDC is here.

  2. The getImageData function returns an object, which contains an array with the byte data of all pixels. Each pixel is described by 4 bytes: r, g, b and a.

    r, g and b are the color components (red, green and blue) and alpha is the opacity. So each pixel uses 4 bytes, and therefore a pixel's data begins at pixel_index * 4.

  3. Yes, it's averaging the values. Because in the next 3 lines r, g and b are all set to that same value, you'll obtain a gray color for each pixel (because the amount of all 3 components are the same).

    So basically, for all pixels this will hold: r === g, g === b and thus also r === b. Colors for which this holds are grayscale (0, 0, 0 being black and 255, 255, 255 being white).

Upvotes: 12

bjornd
bjornd

Reputation: 22943

function grayscale(src){ //Creates a canvas element with a grayscale version of the color image
    //create canvas
    var canvas = document.createElement('canvas');
    //get its context
    var ctx = canvas.getContext('2d');
    //create empty image
    var imgObj = new Image();
    //start to load image from src url
    imgObj.src = src;
    //resize canvas up to size image size
    canvas.width = imgObj.width;
    canvas.height = imgObj.height;
    //draw image on canvas, full canvas API is described here http://www.whatwg.org/specs/web-apps/current-work/multipage/the-canvas-element.html
    ctx.drawImage(imgObj, 0, 0);
    //get array of image pixels
    var imgPixels = ctx.getImageData(0, 0, canvas.width, canvas.height);
    //run through all the pixels
    for(var y = 0; y < imgPixels.height; y++){
        for(var x = 0; x < imgPixels.width; x++){
            //here is x and y are multiplied by 4 because every pixel is four bytes: red, green, blue, alpha
            var i = (y * 4) * imgPixels.width + x * 4; //Why is this multiplied by 4?
            //compute average value for colors, this will convert it to bw
            var avg = (imgPixels.data[i] + imgPixels.data[i + 1] + imgPixels.data[i + 2]) / 3;
            //set values to array
            imgPixels.data[i] = avg; 
            imgPixels.data[i + 1] = avg; 
            imgPixels.data[i + 2] = avg;
        }
    }
    //draw pixels according to computed colors
    ctx.putImageData(imgPixels, 0, 0, 0, 0, imgPixels.width, imgPixels.height); 
    return canvas.toDataURL();
}

In this function coefficient equal to 1/3 are used, however the usually used are: 0.3R + 0.59G + 0.11B (http://gimp-savvy.com/BOOK/index.html?node54.html).

Upvotes: 9

Related Questions