Spaces:
Paused
Paused
| import express from "express" | |
| import fileUpload from "express-fileupload" | |
| import path from "path" | |
| import os from "os" | |
| import fs from "fs" | |
| import archiver from "archiver" | |
| import ffmpeg from "fluent-ffmpeg" | |
| // import { initFolders } from "./initFolders.mts" | |
| import { runColmap, ColmapOptions } from "./colmap.mts" | |
| // initFolders() | |
| declare module 'express-serve-static-core' { | |
| interface Request { | |
| files: any; | |
| } | |
| } | |
| const app = express() | |
| const port = 7860 | |
| const maxActiveRequests = 4 | |
| let activeRequests = 0 | |
| app.use(express.json({limit: '50mb'})) | |
| app.use(express.urlencoded({limit: '50mb', extended: true})) | |
| app.use(fileUpload()) | |
| app.post("/", async (req, res) => { | |
| if (activeRequests >= maxActiveRequests) { | |
| res.status(503).json({message: "Service Unavailable: Max concurrent requests reached. Please try again later"}).end(); | |
| return | |
| } | |
| activeRequests++ | |
| if (!req.files || !req.files.data || req.files.data.mimetype !== 'video/mp4') { | |
| res.status(400).json({error: "Missing or invalid data file in request"}).end() | |
| return | |
| } | |
| let options: ColmapOptions = req.body | |
| let dataFile: fileUpload.UploadedFile = req.files.data | |
| let projectTempDir = path.join(os.tmpdir(), Math.random().toString().slice(2)) | |
| let outputTempDir = path.join(os.tmpdir(), Math.random().toString().slice(2)) | |
| try { | |
| fs.mkdirSync(projectTempDir) | |
| fs.mkdirSync(outputTempDir) | |
| await dataFile.mv(path.join(projectTempDir, dataFile.name)) | |
| let imageFolder = path.join(projectTempDir, 'images'); | |
| fs.mkdirSync(imageFolder) | |
| await new Promise((resolve, reject) => { | |
| ffmpeg(path.join(projectTempDir, dataFile.name)) | |
| .outputOptions('-vf', 'fps=1') // Change this value depending on the number of frames you want from video. | |
| .output(path.join(imageFolder, 'image-%03d.png')) | |
| .on('end', resolve) | |
| .on('error', reject) | |
| .run() | |
| }) | |
| options.projectPath = projectTempDir | |
| options.workspacePath = projectTempDir | |
| options.imagePath = imageFolder | |
| const result = await runColmap(options) | |
| let outputFilePath = path.join(outputTempDir, 'output.zip') | |
| let output = fs.createWriteStream(outputFilePath) | |
| let archive = archiver('zip') | |
| archive.directory(outputTempDir, false) | |
| archive.pipe(output) | |
| await archive.finalize() | |
| res.status(200) | |
| res.download(outputFilePath, 'output.zip', (error) => { | |
| if (!error) fs.rmSync(projectTempDir, {recursive: true, force: true}) | |
| fs.rmSync(outputTempDir, {recursive: true, force: true}) | |
| }) | |
| } catch (error) { | |
| res.status(500).json({ | |
| error: "Couldn't generate pose data", | |
| message: error | |
| }).end() | |
| } finally { | |
| activeRequests-- | |
| } | |
| }); | |
| app.get("/", async (req, res) => { | |
| res.status(200) | |
| res.write(`<html><head></head><body> | |
| Campose API is a micro-service used to generate came pose data from a set of images. | |
| </body></html>`) | |
| res.end() | |
| }) | |
| app.listen(port, () => { console.log(`Open http://localhost:${port}`) }) |