感知识别算法Jetson环境部署测试记录

1. 基础环境配置

sudo apt update
sudo apt install libopenblas-base libopenmpi-dev
conda create --name robot_pytorch21_py38 python=3.8
conda activate robot_pytorch21_py38
pip install numpy
# https://pytorch.org/get-started/previous-versions/
pip install torch==2.1.0 torchvision==0.16.0 torchaudio==2.1.0 --index-url https://download.pytorch.org/whl/cu118

cd ~/model
sudo apt install libjpeg-dev zlib1g-dev
cd torchvision
pip install pillow==10.4.0
pip install flask matplotlib onnx packaging pandas protobuf psutil six
pip install "markdown-it-py[linkify,plugins]" matplotlib pandas pyyaml
pip check
pip install --upgrade $(pip list --outdated --format freeze | cut -d'=' -f1)
python3 setup.py install

2. 非正常热源识别(需要热成像摄像头)

cd efficientvit
pip install cmake
pip install -e ./
pip install -r ../pysenxor/requirements_linux.txt
pip install supervision
sudo chmod 777 /dev/ttyACM0
python scripts/grounding_dino-efficientvit_sam-thermal.py --video assets/light.mp4

3. 生物检测(室外), 图像跟随, 铁丝网识别, 宠物识别, 人员摔倒检测, 火花识别, 积水识别

cd yolov10
pip install -r requirements.txt
pip uninstall setuptools
pip install setuptools
python track.py --video assets/cat.mp4

4. 人脸识别(室内)

cd face
pip install insightface
pip install numpy==1.23.1
pip uninstall opencv-python-headless
pip uninstall opencv-python
pip install opencv-python
sudo apt install libhdf5-dev
pip install h5py==3.10.0
python run.py --mode reg --person-dir C
python run.py --mode ver --video assets/face.mp4

5. 异常噪音检测, 物体摔落检测

cd FunASR
pip install -e ./
pip install modelscope[framework] 
pip install transformers
pip install pyaudio

python scripts/infer.py --audio asserts/demo.wav

6. 仪器仪表识别

cd ultralytics
pip install -e ./
python scripts/infer.py --video assets/far-dark.mp4