Object detection applications to prevent finger injuries in a cutting tool holder environment
research project while in University of Stuttgart
Customized detection area
LED lights up when a finger is detected
Detection speed of about 40fps
Inspired by the following video. (Youtube) The idea is in protecting the safety of the chainsaw operator. When the saw is working, customize the danger zone. If a finger is found in the dangerzone, a signal is output to alert the operator. SawStop | SawStop Hot Dog Demo
- Jetson Nano Developer Kit 4GB
- Micro SD card above 32GB
- USB Webcam
- Breadboard, jumper wires, resistance and LED (for GPIO)
follow the official guide to boot Jetson Nano Getting Started with Jetson Nano Developer Kit
sudo apt-get update
sudo apt-get upgrade
cd ~
gedit .bashrc
Add these three lines at the end, Ctrl+S to save and close gedit
export PATH=/usr/local/cuda/bin:$PATH
export LD_LIBRARY_PATH=/usr/local/cuda/lib64:$LD_LIBRARY_PATH
export CUDA_ROOT=/usr/local/cuda
Activate and reboot
source .bashrc
reboot
3. install pytorch 1.8 and torchvision v0.9.0 via PyTorch for Jetson - version 1.10 now available
wget https://nvidia.box.com/shared/static/p57jwntv436lfrd78inwl7iml6p13fzh.whl -O torch-1.8.0-cp36-cp36m-linux_aarch64.whl
sudo apt-get install python3-pip libopenblas-base libopenmpi-dev
pip3 install Cython
pip3 install numpy torch-1.8.0-cp36-cp36m-linux_aarch64.whl
sudo apt-get install libjpeg-dev zlib1g-dev libpython3-dev libavcodec-dev libavformat-dev libswscale-dev
git clone --branch v0.9.0 https://github.com/pytorch/vision torchvision
cd torchvision
export BUILD_VERSION=0.9.0
python3 setup.py install --user
install dependency
sudo apt-get install liblapack-dev
sudo apt-get install libblas-dev
sudo apt-get install gfortran
install scipy matplotlib pillow pyyaml tensorboard tqdm
pip3 install scipy matplotlib pillow pyyaml tensorboard tqdm
install pillow correctly
pip3 uninstall pillow
pip3 install pillow --no-cache-dir
- install pycuda library
export PATH=/usr/local/cuda-10.2/bin:$PATH
export LD_LIBRARY_PATH=/usr/local/cuda-10.2/lib64:$LD_LIBRARY_PATH
pip3 install pycuda --user
cd ~
git clone https://github.com/newbiehyz/hand_jetsonnano
cd hand_jetsonnano
python3 detect.py # press 'q' to exit