#6060th of 60 in Data and scraping for AI

blenderproc

Python pipeline for rendering synthetic training images with Blender

Stars
3.8k
License
GPL-3.0
Last commit
Jan 2026
Last release
Oct 2024
Language
Python
Stars, last 29 weeks+294
22 Apr 2026: 3,492 stars29 Apr 2026: 3,504 stars6 May 2026: 3,519 stars13 May 2026: 3,531 stars20 May 2026: 3,542 stars27 May 2026: 3,550 stars3 Jun 2026: 3,562 stars10 Jun 2026: 3,570 stars17 Jun 2026: 3,582 stars24 Jun 2026: 3,591 stars1 Jul 2026: 3,596 stars8 Jul 2026: 3,604 stars15 Jul 2026: 3,616 stars22 Jul 2026: 3,624 stars29 Jul 2026: 3,635 stars5 Aug 2026: 3,648 stars12 Aug 2026: 3,660 stars19 Aug 2026: 3,670 stars26 Aug 2026: 3,677 stars2 Sept 2026: 3,686 stars9 Sept 2026: 3,699 stars16 Sept 2026: 3,707 stars23 Sept 2026: 3,722 stars30 Sept 2026: 3,726 stars7 Oct 2026: 3,730 stars14 Oct 2026: 3,743 stars21 Oct 2026: 3,758 stars28 Oct 2026: 3,768 stars4 Nov 2026: 3,774 stars11 Nov 2026: 3,786 stars

3,492 on 22 Apr 20263,786 on 11 Nov 2026

From GitHub's daily star counts.

Checked on GitHub.

Overview

BlenderProc is a Python API and CLI that runs scripts inside Blender's Python environment to build or load 3D scenes, sample object, light and camera poses, and render images. It outputs RGB, stereo, depth, normal and segmentation data, written to .hdf5 containers or COCO and BOP annotations. It is aimed at generating labelled synthetic data for computer vision training.

Who it is for: Computer vision researchers generating synthetic, annotated image datasets

Strengths

  • Loads .obj, .ply, .blend, .fbx, BOP, ShapeNet, Haven and 3D-FRONT scenes
  • Writes .hdf5, COCO and BOP annotations directly from the render step
  • Physics simulation for object placement and sampling of poses, lights and materials
  • Installs with pip; debug subcommand opens the script in the Blender GUI

Weaknesses

  • Scripts must run via the blenderproc CLI, not the normal Python interpreter
  • Breakpoint debugging needs remote debugpy or PyCharm setup inside Blender's Python
  • GPL-3.0 license may constrain use in proprietary code
  • Last tagged release is 2024-10-22, older than recent commits

What it needs

  • Needs Blender
  • Models: obj, ply, blend, fbx, BOP

What changed

From the release notes. No flag means the note does not say so, not that the upgrade is safe.

Also in Data and scraping for AI

See all 60
Also in Data and scraping for AI
RankProjectScore
1doclingPython library that converts PDFs, Office files and more into structured text70.5k stars+1,801 (+2.6%) in 30 daysMIT92 out of 100
2mineruParses PDFs, images and Office files into Markdown and structured JSON Live demo ↗ (opens in a new tab)83.6k stars+2,160 (+2.7%) in 30 dayscustom license91 out of 100
3opendataloader-pdfPDF parser that outputs Markdown, JSON with bounding boxes, and Tagged PDF30.3k stars+705 (+2.4%) in 30 daysApache-2.091 out of 100
#4agent-reachCLI that gives coding agents read access to login-gated and Chinese web platforms111.3k stars+14,960 (+16%) in 30 daysMIT89 out of 100
#5markitdownPython library and CLI that converts office files, PDFs and audio to Markdown195.1k stars+5,652 (+3.0%) in 30 daysMIT87 out of 100