rosbag 使用心得

比较完整的rosbag指令介绍：ROS-wiki

下面写一些我遇到的问题及解决方案：

问题1：

rosbag record -o /home/inin/data/ /occam/image_tiles0 /occam/stitched_image0

我在执行以上指令 录包时会出现丢包的现象，结果如下：

inin@inin:~$ rosbag record -o /home/inin/data/ /occam/image_tiles0 /occam/stitched_image0
[ INFO] [1554260232.649125239]: Subscribing to /occam/image_tiles0
[ INFO] [1554260232.651262266]: Subscribing to /occam/stitched_image0
[ INFO] [1554260232.653578956]: Recording to /home/inin/data/_2019-04-03-10-57-12.bag.
[ WARN] [1554260241.161152134]: rosbag record buffer exceeded.  Dropping oldest queued message.

原因在于我记录的是原始图像，非常大（10s=5g）而rosbag的缓存空间与磁盘写入带宽有限，因此可以从两方面着手。

解决方案1：

法1.提高rosbag的缓存空间：

rosbag record -o /home/inin/data/ -b 4096 /occam/stitched_image0 /occam/image_tiles0

 指令如上所示，rosbag 中加入-b num ,即为将缓存空间设置成num大小，在ROS-wiki也有说明。

 play之后可视化可用：

rosrun image_view image_view image:=/occam/image_tiles0

ros订阅：

class Images
{
public:
    image_transport::Subscriber sub_image;these

    Images()
    {
            ros::NodeHandle node;
            image_transport::ImageTransport it(node);
            sub_image = it.subscribe("/usb_cam/image_raw", 1, &Images::imageCallback, this);
    }

    void imageCallback(const sensor_msgs::ImageConstPtr& msg)
    {
        try
        {
            cv::imshow("view", cv_bridge::toCvShare(msg, "bgr8")->image);
            cv::waitKey(30);
        }
        catch (cv_bridge::Exception& e)
        {
            ROS_ERROR("Could not convert from '%s' to 'bgr8'.", msg->encoding.c_str());

        }

    }
};

int main(int argc, char **argv)
{
    ros::init(argc, argv, "my_node");
    ROS_INFO("my_node running...");
    Images images;
    ros::spin();
    return 0;
}

法2.压缩原始图像大小

rosbag record -o /home/inin/data/ /occam/stitched_image0/compressed /occam/image_tiles0/compressed

存储图像的压缩格式/compressed,可有效减小数据写入大小(1min=1G)，然而压缩后的图像是一种有损压缩对图像质量要求极高的情况慎用，当然一般做视觉计算是无所谓的，相当于把raw格式的图片转换为jpeg格式。

在可视化时指令也有所改变：

rosrun image_view image_view image:=/occam/image_tiles0 compressed

针对压缩图像的ros定义节点可以参考以下：

#include 
#include 
#include 

void imageCallback(const sensor_msgs::CompressedImageConstPtr& msg)
{
  try
  {
    cv::Mat image = cv::imdecode(cv::Mat(msg->data),1);//convert compressed image data to cv::Mat
    cv::imshow("view", image);

    cv::waitKey(10);
  }
  catch (cv_bridge::Exception& e)
  {
    ROS_ERROR("Could not convert to image!");
  }
}

int main(int argc, char **argv)
{
  ros::init(argc, argv, "image_listener");
  ros::NodeHandle nh;
  cv::namedWindow("view");
  cv::startWindowThread();
  ros::Subscriber sub = nh.subscribe("/usb_cam/image_raw/compressed", 1, imageCallback);
  ros::spin();
  cv::destroyWindow("view");
}

问题2：

从一个包含众多ｔｏｐｉｃ的rosbag中分离出我想要的topic:

rosbag filter 2000-01-07-01-07-03.bag split/only-tiles.bag "topic == '/occam/image_tiles0/compressed /InsInfo'"