https://github.com/otl/rostensorflow
TensorFlow ImageNet demo using ROS sensor_msgs/Image
https://github.com/otl/rostensorflow
image-classification image-recognition ros tensorlow
Last synced: 3 months ago
JSON representation
TensorFlow ImageNet demo using ROS sensor_msgs/Image
- Host: GitHub
- URL: https://github.com/otl/rostensorflow
- Owner: OTL
- License: apache-2.0
- Created: 2016-06-25T11:19:02.000Z (almost 9 years ago)
- Default Branch: master
- Last Pushed: 2017-08-03T21:37:59.000Z (almost 8 years ago)
- Last Synced: 2025-03-18T03:51:22.569Z (3 months ago)
- Topics: image-classification, image-recognition, ros, tensorlow
- Language: Python
- Homepage:
- Size: 11.7 KB
- Stars: 61
- Watchers: 10
- Forks: 59
- Open Issues: 3
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
rostensorflow
=====================
- Install TensorFlow (see [tensor flow install guide](https://www.tensorflow.org/install/install_linux))
- Install ROS (see http://wiki.ros.org)
- Install cv-bridge
```bash
$ sudo apt-get install ros-kinetic-cv-bridge ros-kinetic-opencv3
```
- (Optional) Install camera driver (for example, cv_camera)
```bash
$ sudo apt-get install ros-kinetic-cv-camera
```
TensorFlow install note (without GPU)
-------------------------------------------
Please read official guide. This is a only note for me.
```bash
$ sudo apt-get install python-pip python-dev python-virtualenv
$ virtualenv --system-site-packages ~/tensorflow
$ source ~/tensorflow/bin/activate
$ pip install --upgrade tensorflow
```
image_recognition.py
--------------------------------
* publish: /result (std_msgs/String)
* subscribe: /image (sensor_msgs/Image)
How to try
```bash
$ roscore
$ rosrun cv_camera cv_camera_node
$ source ~/tensorflow/bin/activate
$ python image_recognition.py image:=/cv_camera/image_raw
$ rostopic echo /result
```