ROS camera streaming to TensorFlow

tf_ros.py

 def callback(self, image_msg):
        # convert a ROS image message into an cv::Mat using module cv_bridge
        cv_image = self._cv_bridge.imgmsg_to_cv2(image_msg, "bgr8")

        # copy from
        # classify_image.py
        # convert (encode) the image format into streaming data and assign it to 
        # memory cache. It is mainly used for compressing image data format to facilitate
        # network transmission
        image_data = cv2.imencode('.jpg', cv_image)[1].tostring()

        # Creates graph from saved GraphDef.
        softmax_tensor = self._session.graph.get_tensor_by_name('softmax:0')
        predictions = self._session.run(softmax_tensor, {'DecodeJpeg/contents:0': image_data})
        predictions = np.squeeze(predictions)
        
        # Creates node ID --> English string lookup.
        node_lookup = classify_image.NodeLookup()
        top_k = predictions.argsort()[-self.use_top_k:][::-1]
        
        for node_id in top_k:
            human_string = node_lookup.id_to_string(node_id)
            score = predictions[node_id]
            if score > self.score_threshold:
                rospy.loginfo('%s (score = %.5f)' % (human_string, score))
                self._pub.publish(human_string)