glsl-marching-cubes - GitLite

KO | EN

gitlite — search

#react #ai #typescript #go #golang #machine-learning #deep-learning #linux #docker #windows #kubernetes #vue #ios #open-source

glsl-marching-cubes

(★ 149)

WebGL Signed Distance Function Mesher

파일 탐색기

최종 버전 다운로드 (.zip)

- - march.js
- index.js
- main.css
- - jasmine.json
- - bounding-controls.js
  - control-section.js
  - download-controls.js
  - editor-controls.js
  - preview-controls.js
- - calc-potentials.frag
  - shader.vert
- - main.css
- - controls.html
- - split-volume.spec.js
- - march.js
- cubemarch.js
- index.js
- renderer.js
- scene.js
- split-volume.js
- stl-exporter.js
- stl-writer.js
- worker-pool.js
.gitignore
index.html
LICENSE
package-lock.json
package.json
README.md

# 설치 가이드

1. 코드 내려받기

git clone https://github.com/tdhooper/glsl-marching-cubes

깃허브에서 프로젝트 코드 전체를 내 컴퓨터로 내려받습니다.

cd glsl-marching-cubes

방금 내려받은 프로젝트 폴더 안으로 이동합니다.

2. Node.js

쉬움 추천

사전 준비물

Git GitHub에서 프로젝트 코드를 내려받으려면 필요합니다.
Node.js npm 명령어를 쓰려면 Node.js가 설치되어 있어야 합니다. LTS(장기지원) 버전을 추천합니다.

npm install

package.json에 명시된 라이브러리들을 내려받아 설치합니다.

npm start

개발/실행 서버를 켭니다.

✅ 명령어 실행 후 터미널에 나타나는 주소(보통 http://localhost:3000 형태)를 브라우저에서 열어보세요.

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# WebGL Signed Distance Function Mesher

Extract an STL from a glsl distance function

http://tdhooper.github.io/glsl-marching-cubes/

## Limitations

* Large meshes will be saved in parts, you can re-combine them with [MeshLab](http://www.meshlab.net)
* You might get some missing triangles (holes) in the mesh

If you have a Mac, check out [F3](http://www.syedrezaali.com/f3-mac-app/) which does this a LOT better.

## Tools used

* [three.js](http://threejs.org/)
* [stack.gl](http://stack.gl/)
* [TWGL](http://twgljs.org/)
* [FileSaver.js](https://github.com/eligrey/FileSaver.js/)
* [Ractive.js](http://www.ractivejs.org/)

## How does it work?

It follows the basic marching cubes algorithm, except distances for each cube vertex are extracted from the shader first<sup>1</sup>, to reduce the number of GPU/CPU round trips.

The volume to be marched is subdivided into blocks small enough to fit the maximum texture size. As distances are extracted first, any blocks which are entirely inside (negative) or outside (positive) the surface are skipped, saving some time for models with a lot of empty space.

I tried running the entire marching cubes algorithm on the GPU, but the cost and difficulty of getting data back onto the CPU made it far too slow.

The JavaScript marching cubes implementation this is based on<sup>2</sup> runs much faster for simple models. I haven’t compared them for more complex models.

* 1 https://github.com/mikolalysenko/glsl-read-float
* 2 https://github.com/mikolalysenko/isosurface