I will build a computer vision system with opencv on raspberry pi
Embedded Systems and Biomedical Electronics Engineer
Über diesen Service
Need a camera that sees, counts or detects things on its own? I build computer vision systems that run on low-cost edge hardware like the Raspberry Pi, no cloud required.
What I can do for you:
- Object detection and counting (YOLO, MobileNet SSD, TFLite)
- Color, shape and defect inspection with OpenCV
- Tracking, people or vehicle counting, QR/ArUco marker reading
- Camera setup, lighting advice and performance optimization on the Pi
- Integration with GPIO, relays, MQTT or a web dashboard
Why work with me:
I'm an electronics engineering student with a technical degree in industrial automation and hands-on experience with computer vision on Raspberry Pi. I know both the software and the hardware side, so your system works in the real world, not only in a notebook.
How it works:
1. Message me with your use case and a few sample images or videos
2. I confirm feasibility, hardware and the right package
3. You get working code, documentation and a demo
Larger project? I offer custom quotes with milestones.
Programmiersprache:
Python
Tools:
Jupyter-Notizbuch
•
opencv
•
tensorflow
Frameworks:
scikit-learn
•
keras
•
PyTorch
Mein Portfolio
Meine weiteren Dienstleistungen im Bereich Datenwissenschaft & ML
FAQ
Do I need to buy any hardware?
Usually a Raspberry Pi 4 or 5 and a camera module. In the Feasibility Review I recommend the exact board, camera and lens for your use case before you buy anything.
Can it run in real time on a Raspberry Pi?
Yes, for most detection and inspection tasks. I optimize the model (TFLite, smaller input size, frame skipping) and can add an AI accelerator such as Coral or Hailo if you need more speed.
Can you train a model on my own images?
Yes. The Edge Vision System package includes labeling guidance and training a custom model on your dataset. Accuracy depends on how many good, varied images you can provide.

