ROS Resources: Documentation | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange |
2015-03-06 04:33:38 -0500 | answered a question | mysql connection You should add a directory named "cmake-modules" to your project. Inside this directory, create a file with the name "Findmysqlclient.cmake" and copy this content on it: http://www.cmake.org/Wiki/CMakeUserFi... . Then, change your CMakeLists.txt. Replace find_package(MySql REQUIRED) for find_package(mysqlclient REQUIRED) Hope it helps! |
2014-09-16 02:25:53 -0500 | answered a question | rospkg not found in fuerte I think this question has been answered before: |
2014-09-16 02:21:44 -0500 | answered a question | Subscribe to sensor_msgs/PointCloud I think the answer of paulbovbel is the correct one. You just forgot to put a space in this line (between the >> ): distance_sub = nh.subscribe<pcl::pointcloud<pcl::pointxyz>>("/distance_sensors", 1, &SafetyBelt::distanceCallback, this); Change to this: distance_sub = nh.subscribe<pcl::pointcloud<pcl::pointxyz> >("/distance_sensors", 1, &SafetyBelt::distanceCallback, this); |
2014-09-16 02:02:24 -0500 | commented answer | How to use stereo camera for localization? You can also use the stereo_slam package to perform localization using only the stereo images. This package also includes the possibility to reconstruct the world using the pointclouds from the camera. See the videos in the link above ;) |
2013-06-17 06:41:40 -0500 | received badge | ● Nice Answer (source) |
2013-06-04 22:23:44 -0500 | answered a question | How to use stereo camera for localization? You can also use the images of the stereo camera to perform visual odometry. For example, viso2_ros is a ros package that takes stereo image as input and give you the pose of the vehicle. You can integrate both, visual odometry and sonar into a Kalman Filter, but I don't know any package to do this. Robot_pose_ekf integrates visual odometry and imu. Hope this helps... |
2013-05-31 02:24:23 -0500 | answered a question | bfl on groovy, catkin_make errors and compile errors I have exactly the same issue, but I don't have any 'using namespace Eigen;' inside my code. However, I'm importing eigen into my CMakeLists.txt using this: find_package(Eigen REQUIRED) Any ideas? |
2013-04-14 22:20:05 -0500 | commented answer | Unable to subscribe to Openni_launch - dataset doesn't have x-y-z coordinates Yes, this is because you are passing a constant pointer to the setInputCloud() method. You should call this method in that way:PointCloud cloud = *input; sor.setInputCloud(cloud.makeShared()); |
2013-03-24 23:21:16 -0500 | commented question | RGB To Depth mapping Are you using a stereo camera or some RGBD sensor like kinect? |
2013-03-22 03:28:07 -0500 | answered a question | Git: driver cloning doesn't work It seems they have been removed all the projects from their git server (or maybe they are doing maintenance tasks). You can check this here: http://kos.informatik.uni-osnabrueck.de/uos-ros-pkg/?js=1 (link text) where a 404 error is displayed saying: No projects found. I recommend that you contact them directly through its website: http://www.informatik.uni-osnabrueck.de/kbs/staff.html (link text) |
2013-02-27 22:37:15 -0500 | received badge | ● Nice Answer (source) |
2013-02-27 21:24:15 -0500 | commented answer | FOVIS odometry produces NaNs It seems you have sufficient keypoints, and the reference frame is not changing at the moment of NaN. I would suggest you 2 things: Check the way you pass the images to Fovis and try to change some algorithm parameters and see what happens. You can find the parameters in the file stereo-odometry.cpp |
2013-02-27 02:52:49 -0500 | received badge | ● Teacher (source) |
2013-02-27 02:50:38 -0500 | answered a question | FOVIS odometry produces NaNs Hi Spym! It could happen when fovis changes the reference frame. If you launch our wrapper you will see a new ROS topic called /stereo_odometer/info (or the name you set instead). In this topic there is a boolean variable called 'change_reference_frame' which is True when fovis changes the reference frame. Can you, check if you got the NaN when this variable is True? This could help us to isolate the problem... If not, check the other variables in the info message, like 'num_total_keypoints' for example. A small value here could produce NaN |
2013-02-14 03:19:28 -0500 | received badge | ● Supporter (source) |