Computer Graphics & Visualization @ufrj.br
Published
Edited
May 30, 2022
1 fork
21 stars
Canvas primerCanvas primer IISVG IntroMinimal WebGL ProgramWebGL primer with reglShadertoy style drawingTransformation demoModelview DemoModel, View and Projection DemoWebGL CubeCatmull Clark SubdivisionThree.js basicsPhong Illumination modelHalf-edge data structureIterated Function SystemsRaytracingTextured sphereEnvironment mappingPhysical animation with shape matchingShape matching2D GJK and EPA algorithmsOmohundro balltree construction algorithmsPlatonic Solids PlaygroundElliptic billiards 3-periodics in space-time3D LabyrinthBrush drawingBump mappingFractal TerrainModel viewerKeyframe animationTextured planeWallpaper GroupsRay marchingVolume rendering with slice compositingDisplaying 3D RBFs with ray marchingFunções 2D úteisRandom space-filling curvesGPU-IO3D Apollonian Sphere PackingsCircle inversion fractalsApollonian gasket construction by inversionSmooth PolygonDocument image segmentationB-SplinesStencil bufferNoise LoopsSimplex noise exerciserDeformation with RBFsRandom shapes in a gridHobby curvesLane-Riesenfeld subdivision
Simple Adaptive Mosaic Effects
Also listed in…
Paper implementations
portfolio = {
for (;;) {
for (let promise of imagePromises) {
await visibility();
yield await promise;
await Promises.delay(10000);
}
}
}
mutable resetFileInput = null
{
const img = html`<img />`;
img.src = await Files.url(customImg);
mutable image = img;
mutable resetFileInput = mutable resetFileInput + 1;
return `Image ${mutable resetFileInput} Loaded`;
}
//
// Obtain an ImageData object from image
//
imgdata = imageData(image)
//
// Average color and color error for a given imagedata within a given rectangle
//
function getStat(imgdata, x0, y0, w, h) {
const {data, width, height} = imgdata;
const getPixel = (x, y) => {
const offset = (y * width + x) * 4;
return data.slice(offset, offset + 3);
};
let sum = [0, 0, 0];
for (let x = x0; x < x0 + w; ++x) {
for (let y = y0; y < y0 + h; ++y) {
const pix = getPixel(x, y);
sum[0] += pix[0];
sum[1] += pix[1];
sum[2] += pix[2];
}
}
const npix = w * h;
const avg = sum.map((comp) => comp / npix);
let maxErr = 0;
for (let x = x0; x < x0 + w; ++x) {
for (let y = y0; y < y0 + h; ++y) {
const pix = getPixel(x, y);
maxErr = Math.max(
maxErr,
(pix[0] - avg[0]) ** 2 + (pix[1] - avg[1]) ** 2 + (pix[2] - avg[2]) ** 2
);
}
}
return { avg, maxErr };
}
//
// Returns an array of quadrants from building a quadtree on imagedata 'data', such
// that no quadrant has size smaller than minSz and all quadrants have less than the maximum color
// error 'epsilon'.
//
function quadrants(imgdata, epsilon, minSz, maxSz=1024) {
const { data, width, height } = imgdata;
const quads = [];
const visit = (x0, y0, w, h) => {
let { avg, maxErr } = getStat(imgdata, x0, y0, w, h);
if ((Math.max(w, h) <= minSz || maxErr < epsilon) && Math.max(w, h) <= maxSz) {
quads.push({
x0,
y0,
w,
h,
center: [x0 + w / 2, y0 + h / 2],
avg,
maxErr
});
} else {
if (w >= h) {
let sz = 1 << ~~Math.log2(w);
if (sz == w) sz /= 2;
let [l, r, wl, wr] = [x0, x0 + sz, sz, w - sz];
if (h > sz) {
visit(l, y0, wl, sz);
visit(r, y0, wr, sz);
visit(l, y0 + sz, wl, h - sz);
visit(r, y0 + sz, wr, h - sz);
} else {
visit(l, y0, wl, h);
visit(r, y0, wr, h);
}
} else {
let sz = 1 << ~~Math.log2(h);
if (sz == h) sz /= 2;
let [t, b, ht, hb] = [y0, y0 + sz, sz, h - sz];
if (w > sz) {
visit(x0, t, sz, ht);
visit(x0, b, sz, hb);
visit(x0 + sz, t, w - sz, ht);
visit(x0 + sz, b, w - sz, hb);
} else {
visit(x0, t, w, ht);
visit(x0, b, w, hb);
}
}
}
};
visit(0, 0, width, height);
return quads;
}
quads = quadrants(imgdata, sqrtEpsilon ** 2, 1 << logMinSz, 1 << logMaxSz)
{
const del = d3.Delaunay.from(quads.map((q) => q.center));
const { width, height } = imgdata;
const vor = del.voronoi([0, 0, width, height]);
const ctx = DOM.context2d(width, height, 1);
ctx.strokeStyle = "black";
vor.render(ctx);
vor.renderBounds(ctx);
ctx.stroke();
return ctx.canvas;
}
{
const { width, height } = imgdata;
const ctx = DOM.context2d(width, height, 1);
const n = width * height;
const imgdata2 = new ImageData(width, height);
const data = imgdata2.data;
for (let i = 0; i < n; i++) {
const offset = i * 4;
const val = Math.round(luminance[i] * 255);
data[offset] = data[offset + 1] = data[offset + 2] = val;
data[offset + 3] = 255;
}
ctx.putImageData(imgdata2, 0, 0);
return ctx.canvas;
}
luminance = {
const { width, height, data } = imgdata;
const n = width * height;
const lumArray = new Float64Array(n);
for (let i = 0; i < n; i++) {
const offset = i * 4;
lumArray[i] =
(data[offset] * 0.2126 +
data[offset + 1] * 0.7152 +
data[offset + 2] * 0.0722) /
255;
}
return lumArray;
}
{
const { width, height } = imgdata;
const ctx = DOM.context2d(width, height, 1);
const n = width * height;
const imgdata2 = new ImageData(width, height);
const data = imgdata2.data;
for (let i = 0; i < n; i++) {
const offset = i * 4;
const val = Math.round((1 - lumGradient[i]) * 255);
data[offset] = data[offset + 1] = data[offset + 2] = val;
data[offset + 3] = 255;
}
ctx.putImageData(imgdata2, 0, 0);
return ctx.canvas;
}
lumGradient = {
const { width, height } = imgdata;
const n = width * height;
const lumGrad = new Float64Array(n);
const getLum = (x, y) => luminance[y * width + x];
for (let x = 0; x < width; x++) {
let left = Math.max(0, x - 1);
let right = Math.min(width - 1, x + 1);
for (let y = 0; y < height; y++) {
let up = Math.max(0, y - 1);
let down = Math.min(height - 1, y + 1);
lumGrad[y * width + x] = Math.sqrt(
(getLum(right, y) - getLum(left, y)) ** 2 +
(getLum(x, up) - getLum(x, down)) ** 2
);
}
}
return lumGrad;
}
viewof centroidal = {
const { width, height, data } = imgdata;
const n = quads.length;
const points = new Float64Array(n * 2);
quads.forEach(({ center }, i) => {
points[i * 2] = center[0];
points[i * 2 + 1] = center[1];
});
const del = new d3.Delaunay(points);
const vor = del.voronoi([0, 0, width, height]);
const c = new Float64Array(n * 2);
const s = new Float64Array(n);
const getDensity = (x, y) => lumGradient[y * width + x];
const getCentroids = () => {
// Compute the weighted centroid for each Voronoi cell.
c.fill(0);
s.fill(0);
for (let y = 0, i = 0; y < height; ++y) {
for (let x = 0; x < width; ++x) {
const w = getDensity(x, y);
i = del.find(x + 0.5, y + 0.5, i);
s[i] += w;
c[i * 2] += w * (x + 0.5);
c[i * 2 + 1] += w * (y + 0.5);
}
}
for (let i = 0; i < n; ++i) {
const x0 = points[i * 2],
y0 = points[i * 2 + 1];
const x1 = s[i] ? c[i * 2] / s[i] : x0,
y1 = s[i] ? c[i * 2 + 1] / s[i] : y0;
points[i * 2] = x1;
points[i * 2 + 1] = y1;
}
};
const jitter = () => {
for (let i = 0; i < n; ++i) {
// This is necessary to get around a bug in voronoi when centers are aligned in a grid
const x = points[i * 2],
y = points[i * 2 + 1];
points[i * 2] = x + Math.random() * 0.5 - 0.25;
points[i * 2 + 1] = y + Math.random() * 0.5 - 0.25;
}
};
let ctx = DOM.context2d(width, height, 1);
ctx.canvas.value = null;
ctx.strokeStyle = "black";
for (let i = 0; i < lloydIterations; i++) {
getCentroids();
if (i == lloydIterations - 1) jitter();
vor.update();
ctx.clearRect(0, 0, width, height);
ctx.beginPath();
vor.render(ctx);
vor.renderBounds(ctx);
ctx.stroke();
for (let i = 0; i < n; i++) {
ctx.beginPath();
ctx.arc(points[i * 2], points[i * 2 + 1], 1, 0, Math.PI * 2);
ctx.fill();
}
yield ctx.canvas;
}
ctx.canvas.value = vor;
ctx.canvas.dispatchEvent(new CustomEvent("input"));
}
imagePromises = [
FileAttachment("flower2.png").image(),
FileAttachment("butterfly1.png").image(),
FileAttachment("cat1.png").image(),
FileAttachment("flower1.png").image(),
FileAttachment("fruit1.png").image(),
FileAttachment("hanger.png").image(),
FileAttachment("neon.png").image(),
FileAttachment("origami1.png").image(),
FileAttachment("painting1.png").image(),
FileAttachment("portrait1.png").image()
]
import {
image as randomImg,
imageData,
viewof imageConfig
} from "@esperanc/voronoi-mosaics"
import { file } from "@jashkenas/inputs"
One platform to build and deploy the best data apps
Experiment and prototype by building visualizations in live JavaScript notebooks. Collaborate with your team and decide which concepts to build out.
Use Observable Framework to build data apps locally. Use data loaders to build in any language or library, including Python, SQL, and R.
Seamlessly deploy to Observable. Test before you ship, use automatic deploy-on-commit, and ensure your projects are always up-to-date.