SancheZ's profile - activity

Hi, I'm trying to evaluate if a certain point in the map frame is or isn't visible to robot. In order to do so, with the camera angles of view and the robot pose, I can verify if that point is in the camera field of view. But that point, at some views, can be also occluded with something that is on front of it. To verify that I need to check the depth of the pixel where that point would be. If its lower than the depth to that point it is a sign that that point is occluded by something... To perform these transformation from point-> pixel and to verify the field of view I need the intrinsic parameters of my Kinect V1. I don't want to perform a new calibration, since the kinect with the default values seems to be working pretty fine.

When I launch openni I receive the following messages:

[ WARN] [1529322621.477102624]: Camera calibration file /home/some1/.ros/camera_info/rgb_A00362803930039A.yaml not found.
[ WARN] [1529322621.477189197]: Using default parameters for RGB camera calibration.
[ WARN] [1529322621.477267591]: Camera calibration file /home/some1/.ros/camera_info/depth_A00362803930039A.yaml not found.
[ WARN] [1529322621.477340809]: Using default parameters for IR camera calibration.

I don't mind using the default camera parameters, I just need to know them to do my calculations. Do you know where are stored or defined the default kinect v1 parameters that openni is using?

Sorry guys, took about 15 minutes after making this post to found the solution. There's a type of message: sensor_msgs/C

Openni Kinect V1 default parameters Hi, I'm trying to evaluate if a certain point in the map frame is or isn't visible t

Hi, I'm trying to verify if a point in the world is in the field of view of the Kinect that I'm using. I have the coordinates of that point in the world (x,y,z), afterwards I can transform that point to the camera_depth_optical_frame using something like:

transform = self.tf_buffer.lookup_transform('camera_depth_optical_frame', 'map', rospy.Time(0), rospy.Duratio(1.0))
object_on_camera = tf2_geometry_msgs.do_transform_pose(object_on_world, transform) #point on camera_depth_optical_frame

Thanks for your patience, SancheZ

I was able to check if the objects are in the field of view, using the approach that I mentioned above. I project the ob

I was able to check if the objects are in the field of view, using the approach that I mentioned above. I project the ob

I was able to check if the objects are in the field of view, using the approach that I mentioned above. I project the ob

I was able to check if the objects are in the field of view, using the approach that I mentioned above. I project the ob

Verify if world point is in camera's field of view Hi, I'm trying to verify if a point in the world is in the field of v

Verify if world point is in camera's field of view Hi, I'm trying to verify if a point in the world is in the field of v

Verify if world point is in camera's field of view Hi, I'm trying to verify if a point in the world is in the field of v

Openni Launch Hi, I'm working on a object and scene recognition framework that uses deep learning methods. The robot is

Openni Launch Hi, I'm working on a object and scene recognition framework that uses deep learning methods. The robot is

Openni Launch Hi, I'm working on a object and scene recognition framework that uses deep learning methods. The robot is

PointCloud2 read_points unpack_from problem Hey, I'm working on a object recognition framework that uses deep learning m