gaussian.js

(function (imageproc) {
  "use strict";

  /*
   * Apply a Gaussian filter to the input data
   */
  imageproc.gaussianBlur = function (inputData, outputData, size) {
    /* Calculate the sigma value */
    var sigma = size / 4;
    var halfSize = Math.floor(size / 2);

    /* Calculate the row/column matrix */

    var rowMatrix = [];
    for (var i = 0; i < size; i++) {
      rowMatrix[i] =
        Math.exp(-Math.pow(i - halfSize, 2) / (2 * Math.pow(sigma, 2))) /
        (Math.sqrt(2 * Math.PI) * sigma);
    }

    /* Create the kernel */
    var kernel = [];
    var divisor = 0;
    for (var i = 0; i < size; i++) {
      kernel[i] = [];
      for (var j = 0; j < size; j++) {
        kernel[i][j] = rowMatrix[i] * rowMatrix[j];
        divisor += kernel[i][j];
      }
    }

    /***** DO NOT REMOVE - for marking *****/
    console.log("Sigma:", sigma);
    console.log("Kernel:", kernel.toString());
    console.log("Divisor:", divisor);
    /***** DO NOT REMOVE - for marking *****/

    /* Apply the gaussian filter */
    for (var y = 0; y < inputData.height; y++) {
      for (var x = 0; x < inputData.width; x++) {
        var sumR = 0,
          sumG = 0,
          sumB = 0;

        /* Sum the product of the kernel on the pixels */
        for (var j = -halfSize; j <= halfSize; j++) {
          for (var i = -halfSize; i <= halfSize; i++) {
            var pixel = imageproc.getPixel(inputData, x + i, y + j);
            var coeff = kernel[j + halfSize][i + halfSize];

            sumR += pixel.r * coeff;
            sumG += pixel.g * coeff;
            sumB += pixel.b * coeff;
          }
        }

        /* Set the averaged pixel to the output data */
        var i = (x + y * outputData.width) * 4;
        outputData.data[i] = sumR / divisor;
        outputData.data[i + 1] = sumG / divisor;
        outputData.data[i + 2] = sumB / divisor;
      }
    }
  };
})((window.imageproc = window.imageproc || {}));