Tensorflow_in_ROS

I made a ros-node to predict numbers from camera image.
I used Tensorflow tutorial Deep MNIST model(https://www.tensorflow.org/versions/r0.11/tutorials/mnist/pros/index.html)

Before you use this, you must install some programs.

• Install Tensorflow (see https://www.tensorflow.org/versions/r0.9/get_started/os_setup.html)
• Install ROS (see http://wiki.ros.org)
• Install OpenCV(see http://opencv.org/ or https://github.com/opencv/opencv)
• Install cv-bridge, and camera driver (for example, cv_camera)

$ sudo apt-get install ros-indigo-cv-bridge ros-indigo-cv-camera

tensorflow_in_ros_mnist.py

This is main file. This make the ros-node.

• publish: /result (std_msgs/Int16)
• subscribe: /image (sensor_msgs/Image)

model.ckpt,model.ckpt.meta

This is Trained model. Test set accuracy is approximately 99.2%.

How to try

$ roscore
$ python tensorflow_in_ros_mnist.py image:=/cv_camera/image_raw
$ rosrun cv_camera cv_camera_node
$ rostopic echo /result

If you execute this code after launching a camera-node, this code will make error.
But, it is ok because that is just "now loading" error and predict-code will be executed shortly after init-code finished.

Result

You can see ros-node published predicted numbers. Look at this.
Right picture is 9-image from camera.
Left number is published predicted number.

Reference

• Deep MNIST for Experts (https://www.tensorflow.org/versions/r0.11/tutorials/mnist/pros/index.html)
• Variables: Creation, Initialization, Saving, and Loading (https://www.tensorflow.org/versions/r0.11/how_tos/variables/index.html)
• rostensorflow (https://github.com/OTL/rostensorflow)

This software is released under the Apache License2.0, see LICENSE.txt.