Image Record Package for ROS.
- Images shouldn't have to be published in a precise interval.
- If you have a node which computes heavily using image, drawing results on a input image might be delayed.
- Images can be Published in a different fps with output video.
- Images can be published in a variable interval.
- Multiple image topics can be merged in a grid view.
- Log your robot's data in multiple video files can be frustrating, if you have a multiple cameras and multiple calculation nodes.
- Publishing Interval(fps) can be different for camera nodes.
- etc.
This node use opencv to save video, subscribe multiple image-publishing nodes.
-
~output_width(int: default 640) : Output video width.
-
~output_height(int: default 480) : Output video height.
-
~output_fps(int: default 30) : Output video fps(frame per seconds).
-
~output_format(str: default xvid) : Output video format in fourcc format. See FourCC Identifier
-
~output_path(str) : Output Video File Path. eg: /home/ildoonet/Documents/video.mp4
- [timestamp] : will be replaced with real timestamp. eg. grid_[timestamp].mp4 ---> grid_20170626_185926.mp4
-
~output_topic(str) : Broadcast Output Video, if provided. eg. /image_recorder/image_raw
-
~source1(str) : Incoming Video Frame Info in Format(topic,target_x,target_y,target_w,target_h)
- eg: "/cv_camera/image_raw,0,0,320,240"
- Source image will be resized automatically at target_w, target_h
- Resized image will be pasted at the position of (target_x, target_y)
- ~source{n} : n is {1, 2, 3, ...}.
<node name="video_recorder" pkg="video_recorder" type="recorder.py" output="screen" required="true">
<param name="output_width" type="int" value="640" />
<param name="output_height" type="int" value="480" />
<param name="output_path" value="/workdir/result.avi" />
<param name="output_topic" value="$(arg video_topic)" />
<param name="source1" value="/videofile/image_raw,0,0,320,240" /> <!-- left-top -->
<param name="source2" value="/openpose/image_raw,320,0,320,240" /> <!-- right-top -->
<param name="source3" value="/deepdrone/facetrack_img,0,240,320,240" /> <!-- right-bottom -->
</node>
<node name="image_view_grid" pkg="image_view" type="image_view" respawn="false" output="screen" required="true">
<remap from="image" to="$(arg video_topic)"/>
<param name="autosize" value="true" />
</node>
<node name="video_recorder" pkg="video_recorder" type="recorder.py" output="screen" required="true">
<param name="output_width" type="int" value="640" />
<param name="output_height" type="int" value="480" />
<param name="output_path" value="/workdir/result.avi" />
<param name="output_topic" value="$(arg video_topic)" />
<param name="source1" value="/videofile/image_raw,0,0,640,480" /> <!-- left-top with big size(640x480) -->
<param name="source2" value="/openpose/image_raw,640,0,320,240" /> <!-- right-top (320x240) -->
<param name="source3" value="/deepdrone/facetrack_img,640,240,320,240" /> <!-- right-bottom (320x240) -->
<param name="source4" value="/deepdrone/tracking_img,0,480,320,240" /> <!-- left-bottom (320x240) -->
</node>
<node name="image_view_grid" pkg="image_view" type="image_view" respawn="false" output="screen" required="true">
<remap from="image" to="$(arg video_topic)"/>
<param name="autosize" value="true" />
</node>
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent