Embedded Linux ▪ Imaging Pipelines ▪ libcamera ▪ Linux Kernel Media ▪ RKISP1 ▪ Edge ML
I build end‑to‑end camera systems: from sensor bring‑up → ISP tuning → libcamera integration → real‑time pipelines → edge ML → deployment on embedded Linux.
Currently transitioning into a more open‑source focused role with contributions to:
- libcamera (RKISP1 pipeline, tooling, docs)
- Linux media subsystem (rkisp1, V4L2, device tree improvements, sensor drivers)
- Calibration, optics & controlled illumination
- Mainline Linux rkisp1 pipeline for the i.MX8MP
- Calibrated imaging pipeline with libcamera + ISP metadata
- Edge ML (ONNX / TFLite) integrated into a real-time embedded camera system
- Open‑source contributions in imaging, Linux kernel media, and embedded Linux
- Technical blog on: https://felixsteinberger.dev (coming soon)
- Mainline Linux
- systemd ▪ device trees
- SPI ▪ I2C ▪ UART ▪ MIPI CSI‑2
- STM32
- libcamera pipeline design
- V4L2 / media controller
- Sensor bring-up (MT925, AR0521, others)
- ISP concepts: AE, AWB, LSC, demosaic, NR
- Optics & controlled illumination
- C/C++ (embedded focus)
- GStreamer pipelines
- Performance profiling (perf, strace, valgrind)
- ONNX Runtime ▪ TFLite
- Model quantization (INT8)
- Real-time inference on constrained hardware
- ✔ Contribute upstream to libcamera
- ✔ Submit patches to Linux media subsystem
- ✔ Publish deep-dive blog posts on imaging pipelines
- ✔ Release open-source tools for calibration & metadata exploration
- ✔ Build a purpose-driven ML imaging project (medtech / sustainability)
I write about imaging, embedded Linux, calibration workflows, and edge ML on:
➡️ https://felixsteinberger.dev (launching soon)
Topics planned:
- Mainline rkisp1 + libcamera deep dive
- Building reliable camera systems
- Device tree pitfalls for MIPI cameras
- From ISP metadata to ML pre-processing
- Real-world calibration (illumination, lenses, geometry)
- libcamera improvements (tools, RKISP1, docs)
- Linux kernel media (V4L2, DT, ISP)
- Imaging pipelines for industrial/medtech/sustainability
- Embedded Linux camera devices
- Edge ML applied to real-time vision
- Future freelance consulting (embedded imaging + Linux camera stack architecture)
- Email: felix@felixsteinberger.dev
- LinkedIn: https://www.linkedin.com/in/felix-steinberger-475317399/
- Blog: https://felixsteinberger.dev
Let’s build better embedded imaging systems together!