image_classification.md
March 29, 2018 ยท View on GitHub
Classification for Image
This project supports multiple CNN models for classification. Please make sure you have already set up environment and installed this project correctly. You can refer to the following links for your interested models then.
1 CNN Models
2 Other Arguments
1 CNN Models
Compile NCS graph.
cd /opt/movidius/ncappzoo/caffe/AlexNet
make
Launch object classification service.
cd ~/ros2_ws
echo "param_file: alexnet.yaml" > src/ros2_intel_movidius_ncs/movidius_ncs_launch/config/default.yaml
ament build
source install/local_setup.bash
launch `ros2 pkg prefix movidius_ncs_launch`/share/movidius_ncs_launch/launch/ncs_image_launch.py
Run the example application with an absolute path of an image on another console. For example,
ros2 run movidius_ncs_example movidius_ncs_example_image_classification /opt/movidius/ncappzoo/data/images/cat.jpg
Compile NCS graph.
cd /opt/movidius/ncappzoo/caffe/GoogleNet
make
Launch object classification service.
cd ~/ros2_ws
echo "param_file: googlenet.yaml" > src/ros2_intel_movidius_ncs/movidius_ncs_launch/config/default.yaml
ament build
source install/local_setup.bash
launch `ros2 pkg prefix movidius_ncs_launch`/share/movidius_ncs_launch/launch/ncs_image_launch.py
Run the example application with an absolute path of an image on another console. For example,
ros2 run movidius_ncs_example movidius_ncs_example_image_classification /opt/movidius/ncappzoo/data/images/cat.jpg
Compile NCS graph.
cd /opt/movidius/ncappzoo/caffe/SqueezeNet
make
Launch object classification service.
cd ~/ros2_ws
echo "param_file: squeezenet.yaml" > src/ros2_intel_movidius_ncs/movidius_ncs_launch/config/default.yaml
ament build
source install/local_setup.bash
launch `ros2 pkg prefix movidius_ncs_launch`/share/movidius_ncs_launch/launch/ncs_image_launch.py
Run the example application with an absolute path of an image on another console. For example,
ros2 run movidius_ncs_example movidius_ncs_example_image_classification /opt/movidius/ncappzoo/data/images/cat.jpg
Compile NCS graph.
cd /opt/movidius/ncappzoo/tensorflow/inception_v1
make
Launch object classification service.
cd ~/ros2_ws
echo "param_file: inception_v1.yaml" > src/ros2_intel_movidius_ncs/movidius_ncs_launch/config/default.yaml
ament build
source install/local_setup.bash
launch `ros2 pkg prefix movidius_ncs_launch`/share/movidius_ncs_launch/launch/ncs_image_launch.py
Run the example application with an absolute path of an image on another console. For example,
ros2 run movidius_ncs_example movidius_ncs_example_image_classification /opt/movidius/ncappzoo/data/images/cat.jpg
Compile NCS graph.
cd /opt/movidius/ncappzoo/tensorflow/inception_v2
make
Launch object classification service.
cd ~/ros2_ws
echo "param_file: inception_v2.yaml" > src/ros2_intel_movidius_ncs/movidius_ncs_launch/config/default.yaml
ament build
source install/local_setup.bash
launch `ros2 pkg prefix movidius_ncs_launch`/share/movidius_ncs_launch/launch/ncs_image_launch.py
Run the example application with an absolute path of an image on another console. For example,
ros2 run movidius_ncs_example movidius_ncs_example_image_classification /opt/movidius/ncappzoo/data/images/cat.jpg
Compile NCS graph.
cd /opt/movidius/ncappzoo/tensorflow/inception_v3
make
Launch object classification service.
cd ~/ros2_ws
echo "param_file: inception_v3.yaml" > src/ros2_intel_movidius_ncs/movidius_ncs_launch/config/default.yaml
ament build
source install/local_setup.bash
launch `ros2 pkg prefix movidius_ncs_launch`/share/movidius_ncs_launch/launch/ncs_image_launch.py
Run the example application with an absolute path of an image on another console. For example,
ros2 run movidius_ncs_example movidius_ncs_example_image_classification /opt/movidius/ncappzoo/data/images/cat.jpg
Compile NCS graph.
cd /opt/movidius/ncappzoo/tensorflow/inception_v4
make
Launch object classification service.
cd ~/ros2_ws
echo "param_file: inception_v4.yaml" > src/ros2_intel_movidius_ncs/movidius_ncs_launch/config/default.yaml
ament build
source install/local_setup.bash
launch `ros2 pkg prefix movidius_ncs_launch`/share/movidius_ncs_launch/launch/ncs_image_launch.py
Run the example application with an absolute path of an image on another console. For example,
ros2 run movidius_ncs_example movidius_ncs_example_image_classification /opt/movidius/ncappzoo/data/images/cat.jpg
Compile NCS graph.
cd /opt/movidius/ncappzoo/tensorflow/mobilenets
make
Launch object classification service.
cd ~/ros2_ws
echo "param_file: mobilenet.yaml" > src/ros2_intel_movidius_ncs/movidius_ncs_launch/config/default.yaml
ament build
source install/local_setup.bash
launch `ros2 pkg prefix movidius_ncs_launch`/share/movidius_ncs_launch/launch/ncs_image_launch.py
Run the example application with an absolute path of an image on another console. For example,
ros2 run movidius_ncs_example movidius_ncs_example_image_classification /opt/movidius/ncappzoo/data/images/cat.jpg
2 Other Arguments
| Arguments | Description | Default Value | Valid Values |
|---|---|---|---|
| device_index | ncs device index | 0 | 0~N-1(N is the maximum number of inserted NCS devices) |
| log_level | ncs log level | 0 | 0:Nothing / 1:Errors / 2:Verbose |
| cnn_type | indicate different cnn models | googlenet | alexnet / googlenet / squeezenet / inception_v1 / inception_v2 / inception_v3 / inception_v4 / mobilenet |
| top_n | the number of results to be shown, only valid for classification | 3 | 0~5 |