jsk-ros-pkg/jsk_recognition

JSK perception ROS packages

jsk_recognition's Issues

groovy/catkinでjsk_pcl_rosがコンパイルできない

package.xmlにpclを含んでいない

• hydroではpclは含まない
• groovyではpclを含むべき (未対応)

LightWeight throttle for nodelet

original throttle program of nodelet_topic_tools is pretty heavy because of queue length.

nodelet_topic_tools throttle utilizes 10 length queue, we can use 1 length queue and
put sleep to drop messages.

release

please let me know if you're ready to release

DelayPointCloudは使っているのか

DelayPointCloudをまだ使っているか調べて、そのようなプログラムがあったらVoxelGridManagerに置き換える.

@tarukosu

particleFilterTrackingのパラメータをdynamic_reconfigureで設定できるようにする

euclidean_segmentation.launch failed

ちょっとデバッグが困難なので分かる人におしえてもらいたいのですが、
以下のようなエラーでいくつかのノードレットが死んでしまいます。
何が原因として考えられるでしょうか？

[http://hrp2008v:10008] leus@radish:~$ roslaunch jsk_pcl_ros euclidean_segmentation.launch input_point:=/camera_remote/depth_registered/points manager:=/camera/camera_nodelet_manager
... logging to /home/leus/.ros/log/5ecd514c-69a5-11e4-b183-0090050085a9/roslaunch-radish-32476.log
Checking log directory for disk usage. This may take awhile.
Press Ctrl-C to interrupt
Done checking log file disk usage. Usage is <1GB.

started roslaunch server http://192.168.101.167:37036/

SUMMARY
========

CLEAR PARAMETERS
 * /pcl_nodelet/centroid_publisher00/
 * /pcl_nodelet/centroid_publisher01/
 * /pcl_nodelet/centroid_publisher02/
 * /pcl_nodelet/centroid_publisher03/
 * /pcl_nodelet/centroid_publisher04/
 * /pcl_nodelet/centroid_publisher05/
 * /pcl_nodelet/centroid_publisher06/
 * /pcl_nodelet/centroid_publisher07/
 * /pcl_nodelet/centroid_publisher08/
 * /pcl_nodelet/centroid_publisher09/
 * /pcl_nodelet/cluster_decomposer/
 * /pcl_nodelet/clustering/
 * /pcl_nodelet/voxelgrid/

PARAMETERS
 * /pcl_nodelet/centroid_publisher00/frame
 * /pcl_nodelet/centroid_publisher01/frame
 * /pcl_nodelet/centroid_publisher02/frame
 * /pcl_nodelet/centroid_publisher03/frame
 * /pcl_nodelet/centroid_publisher04/frame
 * /pcl_nodelet/centroid_publisher05/frame
 * /pcl_nodelet/centroid_publisher06/frame
 * /pcl_nodelet/centroid_publisher07/frame
 * /pcl_nodelet/centroid_publisher08/frame
 * /pcl_nodelet/centroid_publisher09/frame
 * /pcl_nodelet/clustering/tolerance
 * /pcl_nodelet/voxelgrid/filter_field_name
 * /pcl_nodelet/voxelgrid/filter_limit_max
 * /pcl_nodelet/voxelgrid/filter_limit_min
 * /pcl_nodelet/voxelgrid/leaf_size
 * /rosdistro
 * /rosversion

NODES
  /pcl_nodelet/
    centroid_publisher00 (nodelet/nodelet)
    centroid_publisher01 (nodelet/nodelet)
    centroid_publisher02 (nodelet/nodelet)
    centroid_publisher03 (nodelet/nodelet)
    centroid_publisher04 (nodelet/nodelet)
    centroid_publisher05 (nodelet/nodelet)
    centroid_publisher06 (nodelet/nodelet)
    centroid_publisher07 (nodelet/nodelet)
    centroid_publisher08 (nodelet/nodelet)
    centroid_publisher09 (nodelet/nodelet)
    cluster_decomposer (nodelet/nodelet)
    clustering (nodelet/nodelet)
    voxelgrid (nodelet/nodelet)

ROS_MASTER_URI=http://hrp2008v:10008

core service [/rosout] found
process[pcl_nodelet/voxelgrid-1]: started with pid [32485]
[ INFO] [1415766864.584009813]: Loading nodelet /pcl_nodelet/voxelgrid of type pcl/VoxelGrid to manager /camera/camera_nodelet_manager with the following remappings:
[ INFO] [1415766864.584077680]: /pcl_nodelet/voxelgrid/input -> /camera_remote/depth_registered/points
process[pcl_nodelet/clustering-2]: started with pid [32491]
[ INFO] [1415766865.939797311]: Loading nodelet /pcl_nodelet/clustering of type jsk_pcl/EuclideanClustering to manager /camera/camera_nodelet_manager with the following remappings:
[ INFO] [1415766865.939868013]: /pcl_nodelet/clustering/input -> /pcl_nodelet/voxelgrid/output
process[pcl_nodelet/cluster_decomposer-3]: started with pid [32496]
[ INFO] [1415766867.228423668]: Loading nodelet /pcl_nodelet/cluster_decomposer of type jsk_pcl/ClusterPointIndicesDecomposerZAxis to manager /camera/camera_nodelet_manager with the following remappings:
[ INFO] [1415766867.228492208]: /pcl_nodelet/cluster_decomposer/input -> /pcl_nodelet/voxelgrid/output
[ INFO] [1415766867.228509544]: /pcl_nodelet/cluster_decomposer/target -> /pcl_nodelet/clustering/output
process[pcl_nodelet/centroid_publisher00-4]: started with pid [32501]
[ INFO] [1415766868.524759466]: Loading nodelet /pcl_nodelet/centroid_publisher00 of type jsk_pcl/CentroidPublisher to manager /camera/camera_nodelet_manager with the following remappings:
[ INFO] [1415766868.524826410]: /pcl_nodelet/centroid_publisher00/input -> /pcl_nodelet/cluster_decomposer/output00
process[pcl_nodelet/centroid_publisher01-5]: started with pid [32506]
[ INFO] [1415766869.831416171]: Loading nodelet /pcl_nodelet/centroid_publisher01 of type jsk_pcl/CentroidPublisher to manager /camera/camera_nodelet_manager with the following remappings:
[ INFO] [1415766869.831491866]: /pcl_nodelet/centroid_publisher01/input -> /pcl_nodelet/cluster_decomposer/output01
process[pcl_nodelet/centroid_publisher02-6]: started with pid [32511]
[ INFO] [1415766871.171945616]: Loading nodelet /pcl_nodelet/centroid_publisher02 of type jsk_pcl/CentroidPublisher to manager /camera/camera_nodelet_manager with the following remappings:
[ INFO] [1415766871.172023889]: /pcl_nodelet/centroid_publisher02/input -> /pcl_nodelet/cluster_decomposer/output02
[FATAL] [1415766871.289157285]: Service call failed!
[pcl_nodelet/clustering-2] process has died [pid 32491, exit code 255, cmd /opt/ros/hydro/lib/nodelet/nodelet load jsk_pcl/EuclideanClustering /camera/camera_nodelet_manager ~input:=/pcl_nodelet/voxelgrid/output __name:=clustering __log:=/home/leus/.ros/log/5ecd514c-69a5-11e4-b183-0090050085a9/pcl_nodelet-clustering-2.log].
log file: /home/leus/.ros/log/5ecd514c-69a5-11e4-b183-0090050085a9/pcl_nodelet-clustering-2*.log
process[pcl_nodelet/centroid_publisher03-7]: started with pid [32516]
[ INFO] [1415766872.511880296]: Loading nodelet /pcl_nodelet/centroid_publisher03 of type jsk_pcl/CentroidPublisher to manager /camera/camera_nodelet_manager with the following remappings:
[ INFO] [1415766872.511946183]: /pcl_nodelet/centroid_publisher03/input -> /pcl_nodelet/cluster_decomposer/output03
process[pcl_nodelet/centroid_publisher04-8]: started with pid [32521]
[ INFO] [1415766873.824439940]: Loading nodelet /pcl_nodelet/centroid_publisher04 of type jsk_pcl/CentroidPublisher to manager /camera/camera_nodelet_manager with the following remappings:
[ INFO] [1415766873.824520879]: /pcl_nodelet/centroid_publisher04/input -> /pcl_nodelet/cluster_decomposer/output04
[FATAL] [1415766874.357654795]: Service call failed!
[pcl_nodelet/cluster_decomposer-3] process has died [pid 32496, exit code 255, cmd /opt/ros/hydro/lib/nodelet/nodelet load jsk_pcl/ClusterPointIndicesDecomposerZAxis /camera/camera_nodelet_manager ~input:=/pcl_nodelet/voxelgrid/output ~target:=/pcl_nodelet/clustering/output __name:=cluster_decomposer __log:=/home/leus/.ros/log/5ecd514c-69a5-11e4-b183-0090050085a9/pcl_nodelet-cluster_decomposer-3.log].
log file: /home/leus/.ros/log/5ecd514c-69a5-11e4-b183-0090050085a9/pcl_nodelet-cluster_decomposer-3*.log
process[pcl_nodelet/centroid_publisher05-9]: started with pid [32526]
[ INFO] [1415766875.103614999]: Loading nodelet /pcl_nodelet/centroid_publisher05 of type jsk_pcl/CentroidPublisher to manager /camera/camera_nodelet_manager with the following remappings:
[ INFO] [1415766875.103680220]: /pcl_nodelet/centroid_publisher05/input -> /pcl_nodelet/cluster_decomposer/output05
process[pcl_nodelet/centroid_publisher06-10]: started with pid [32531]
[ INFO] [1415766876.418460039]: Loading nodelet /pcl_nodelet/centroid_publisher06 of type jsk_pcl/CentroidPublisher to manager /camera/camera_nodelet_manager with the following remappings:
[ INFO] [1415766876.418531074]: /pcl_nodelet/centroid_publisher06/input -> /pcl_nodelet/cluster_decomposer/output06
process[pcl_nodelet/centroid_publisher07-11]: started with pid [32536]
[ INFO] [1415766877.769507429]: Loading nodelet /pcl_nodelet/centroid_publisher07 of type jsk_pcl/CentroidPublisher to manager /camera/camera_nodelet_manager with the following remappings:
[ INFO] [1415766877.769574400]: /pcl_nodelet/centroid_publisher07/input -> /pcl_nodelet/cluster_decomposer/output07
[FATAL] [1415766878.718338138]: Service call failed!
process[pcl_nodelet/centroid_publisher08-12]: started with pid [32541]
[ INFO] [1415766879.073877191]: Loading nodelet /pcl_nodelet/centroid_publisher08 of type jsk_pcl/CentroidPublisher to manager /camera/camera_nodelet_manager with the following remappings:
[ INFO] [1415766879.073966707]: /pcl_nodelet/centroid_publisher08/input -> /pcl_nodelet/cluster_decomposer/output08
[pcl_nodelet/centroid_publisher00-4] process has died [pid 32501, exit code 255, cmd /opt/ros/hydro/lib/nodelet/nodelet load jsk_pcl/CentroidPublisher /camera/camera_nodelet_manager ~input:=/pcl_nodelet/cluster_decomposer/output00 __name:=centroid_publisher00 __log:=/home/leus/.ros/log/5ecd514c-69a5-11e4-b183-0090050085a9/pcl_nodelet-centroid_publisher00-4.log].
log file: /home/leus/.ros/log/5ecd514c-69a5-11e4-b183-0090050085a9/pcl_nodelet-centroid_publisher00-4*.log
process[pcl_nodelet/centroid_publisher09-13]: started with pid [32546]
[ INFO] [1415766880.439712590]: Loading nodelet /pcl_nodelet/centroid_publisher09 of type jsk_pcl/CentroidPublisher to manager /camera/camera_nodelet_manager with the following remappings:
[ INFO] [1415766880.439927973]: /pcl_nodelet/centroid_publisher09/input -> /pcl_nodelet/cluster_decomposer/output09
[FATAL] [1415766883.946575941]: Service call failed!
[pcl_nodelet/centroid_publisher01-5] process has died [pid 32506, exit code 255, cmd /opt/ros/hydro/lib/nodelet/nodelet load jsk_pcl/CentroidPublisher /camera/camera_nodelet_manager ~input:=/pcl_nodelet/cluster_decomposer/output01 __name:=centroid_publisher01 __log:=/home/leus/.ros/log/5ecd514c-69a5-11e4-b183-0090050085a9/pcl_nodelet-centroid_publisher01-5.log].
log file: /home/leus/.ros/log/5ecd514c-69a5-11e4-b183-0090050085a9/pcl_nodelet-centroid_publisher01-5*.log
av^C[pcl_nodelet/centroid_publisher09-13] killing on exit
[pcl_nodelet/centroid_publisher08-12] killing on exit
 [pcl_nodelet/centroid_publisher07-11] killing on exit
[pcl_nodelet/centroid_publisher06-10] killing on exit
[pcl_nodelet/centroid_publisher05-9] killing on exit
[pcl_nodelet/centroid_publisher04-8] killing on exit
[pcl_nodelet/centroid_publisher03-7] killing on exit
[pcl_nodelet/centroid_publisher02-6] killing on exit
[pcl_nodelet/voxelgrid-1] killing on exit
[ INFO] [1415766889.570565029]: Unloading nodelet /pcl_nodelet/voxelgrid from manager /camera/camera_nodelet_manager
[pcl_nodelet/voxelgrid-1] escalating to SIGTERM
shutting down processing monitor...
... shutting down processing monitor complete
done

Is jsk_freenect2 not released as .deb on purpose

Someone is wondering here.

move image_shrink to jsk_recognition

Let's move image_shrink package to jsk_recognition.
Is there any problem?

I believe that package should be installed as default.

[jsk_pcl_ros] VoxelGrid{Manager,Decoder} should use proper message

because of a historical reason, it hacks sensor_msgs/PointCloud2 frame_id.
we should use more proper topic type

dependancy to pcl in jsk_perception

In hydro, rosdep in jsk_perception cause error on dependancy to pcl in manifest.xml because there is not ros-hydro-pcl. (When I comment out , rosdep successfully completed.)
How do I make both hydro and groovy successfully through rosdep?

[jsk_pcl_ros] special nodes to run single nodelet should be auto-generated

We can generate cpp codes to run single nodelet such as pointcloud_screenpoint.cpp.

iai_kinect2は、ubuntu12.04に使うできだ

The source code in github of IAI_kinect2 and Libfreenect2 are changing everyday.
Any one who wants to use the library plz carefully choose a stable commit
and also a small adjustment should be placed.

1##. (Firstly U have to check whether ur GPU card is working or not)

2##. git clone https://github.com/code-iai/iai_kinect2 to ur catkin workspace and also

install the libfreenect2(https://github.com/OpenKinect/libfreenect2) to any other place.
I'm using commit 41d35b0f0cb812bd2e0a4f35b676becb5e1d8b26 Fri Oct 31 for iai_kinect2
and 22eeda006eedc03f02815809bb2a7d4ce2351d1f Thu Oct 9 for libfreenect2. U can update of course.

3##. The packages u need to download(

automake
libudev-dev
xorg-dev
libturbojpeg
ocl-icd-lib*
ocl-icd-open*(need add trusty repository to ur sourcelist)
opencl-headers
....cant remember... )

4##. There may be some link error, locate the library file and ln -s to the require path (basically in path /usr/lib/ /usr/lib/x86*/)

5##. Sometimes CMakeList also need to fix.

6##. Firstly compile libfreenect2, then iai_kinect2. I hope I could create a pull, but I am not familiar with this yet...

7##. Good Luck!

jsk_pcl_rosで不必要なnodeletを整理する

Pointcloud of stereo camera from multisense is not organized

I've already reported https://bitbucket.org/crl/multisense_ros/issue/14/pointcloud-from-stereo-camera-is-not

Publishing tracked object TF transformation in jsk_pcl_ros tracker

Make jsk_pcl_ros tracker publish the TF transformation of the tracked object.

organized_multi_plane_segmentation.launchに特定物体認識を加えたい

jsk_recognitionのorganized_multi_plane_segmentation.launchで物体に対して、interactive_markerでタッチできるので、
そのときに送られてくる点群を登録して、再び見えたときに、それだよと教えてくれる
機能を追加したいです。

pclを使った認識がチュートリアル
http://pointclouds.org/documentation/tutorials/correspondence_grouping.php#correspondence-grouping
にあるので、それを組み込めばいいんじゃないかと思ってます。

あるいは、いい感じの方法や既存のプログラムなどがありますでしょうか

Remove jsk_libfreenect2?

iai_kinect2 seems to be better than jsk_libfreenect2.
Should we remove jsk_libfreenect2?

Does anyone try iai_kinect2?

認識の並列化について

冷蔵庫デモの実装上の相談です。
以前自分が書いた、認識のみ別個で走らせるコードは、立ち上げると常に認識結果を更新するようなコードになっており、かなり重いです。そのため、冷蔵庫デモを実際に行う際に立ち上げるlaunchでは、この部分のコードを採用しておらず、認識が必要になった時にObjectDetectionをsubscribeし始めるようになっている(jsk_perception/euslisp/detection-interface.l)ので、認識結果が必要な時のみ認識処理を回すようなコードとなっています。
認識結果を関数型ではなくactionlibで受け取るように変更したいので、認識処理を常に回す形に移行したいのですが、処理の重たさ故に冷蔵庫デモに支障を来すでしょうか？
必要ない時処理速度を落とすといった、賢い実装をした上でdemoのプログラムとつなげるべきでしょうか？

具体的には、自分が書いた認識のみ別個で走らせるコード(jsk_perception/euslisp/kalman-filtered-objectdetction-marker.l)にactionserverとしての機能を持たせ、デモの中で認識が必要になった時に、その最新の認識結果を元のコード(jsk_perception/euslisp/detection-interface.l)がactionclientとして要求する形にしようとしています。

refactor filter classes

Filter関連が複雑すぎるので、もっと簡単な形にして管理しやすくする。

農園の認識に関するパッケージを置きたい

jsk_tilt_laserを使って取得した農園の中で
農作物が植えられた畝と畝の間の通路の平面を検出したり
検出した平面に主成分分析を適用して移動の進行方向を示したり
通路幅を自動計測して表示するパッケージを作ったので
jsk_pcl_rosかどこかに置きたいと思っています

名前はjsk_gardenとかにしようと思います

libsiftfastについて

　point_pose_extractor_sample.launchを使う際に,imagesiftパッケージのimagesiftを使っています。この中で、libsiftfast.hの関数を使っています。
　siftのパラメーターを変えて使いたいと思っているのですが、これだと難しいです。
opencvのsiftを使うようにできないでしょうか？

depth imageのscale/ROI指定がほしい

depth imageでscaleやROIを設定すると、適したsensor_msgs/Imageとsensor_msgs/CameraInfoが出てくるものがほしいです。

OrganizedなPointCloudのほうが高速な処理が期待できるためです

release as deb package

I'll plan to release jsk_recognition as dep package.

Currently I'm waiting for

• jsk_perception #23

if you have another package to be released let me know.

NOTE: I'll plan to release package very frequently

Fail to load jsk_pcl/ImageRotateNodelet

When loading jsk_pcl/ImageRotateNodlet, I received following errors.

[ERROR] [1398840921.266327158]: Service call failed: service [/openni_nodelet_manager/load_nodelet] responded with an error: boost::lock_error
[FATAL] [1398840921.266405741]: Service call failed!
[l_forearm_cam/lhand_resized_rotated-7] process has died [pid 8147, exit code 255, cmd /opt/ros/groovy/lib/nodelet/nodelet load jsk_pcl/ImageRotateNodelet /openni_nodelet_manager image:=image_rect __name:=lhand_resized_rotated __log:=/home/applications/.ros/log/143be0d2-d010-11e3-9ea1-001517ebc305/l_forearm_cam-lhand_resized_rotated-7.log].
log file: /home/applications/.ros/log/143be0d2-d010-11e3-9ea1-001517ebc305/l_forearm_cam-lhand_resized_rotated-7*.log
[FATAL] [1398840921.642748949]: Service call failed!

Segmentation fault in sparce_image_decorder

Sparce_image_decorder is sometimes killed by segmentation fault.

(gdb) bt
#0  0x00007ffff6d67425 in __GI_raise (sig=<optimized out>)
    at ../nptl/sysdeps/unix/sysv/linux/raise.c:64
#1  0x00007ffff6d6ab8b in __GI_abort () at abort.c:91
#2  0x00007ffff6d600ee in __assert_fail_base (fmt=<optimized out>, 
    assertion=0x7fffec311df6 "px != 0", 
    file=0x7fffec3121d0 "/usr/include/boost/smart_ptr/shared_ptr.hpp", 
    line=<optimized out>, function=<optimized out>) at assert.c:94
#3  0x00007ffff6d60192 in __GI___assert_fail (
    assertion=0x7fffec311df6 "px != 0", 
    file=0x7fffec3121d0 "/usr/include/boost/smart_ptr/shared_ptr.hpp", 
    line=418, 
    function=0x7fffec316cc0 "T* boost::shared_ptr<T>::operator->() const [with T = sensor_msgs::Image_<std::allocator<void> >]") at assert.c:103
#4  0x00007fffec2ed718 in boost::shared_ptr<sensor_msgs::Image_<std::allocator<void> > >::operator-> (this=<optimized out>)
    at /usr/include/boost/smart_ptr/shared_ptr.hpp:418
#5  0x00007fffec30e67a in operator-> (this=<optimized out>)
    at /usr/include/boost/smart_ptr/shared_ptr.hpp:418
#6  jsk_perception::SparseImageDecoder::do_work (this=0x613170, msg=..., 
    input_frame_from_msg=...)
    at /home/furuta/ros/groovy/jsk-ros-pkg/jsk_recognition/jsk_perception/src/sparse_image_decoder.cpp:63
#7  0x00007fffec30e936 in jsk_perception::SparseImageDecoder::imageCallback (
---Type <return> to continue, or q <return> to quit---
    this=0x613170, msg=...)
    at /home/furuta/ros/groovy/jsk-ros-pkg/jsk_recognition/jsk_perception/src/sparse_image_decoder.cpp:28
#8  0x00007fffec31059a in operator() (a0=..., this=<optimized out>)
    at /usr/include/boost/function/function_template.hpp:1013
#9  boost::detail::function::void_function_obj_invoker1<boost::function<void (boost::shared_ptr<jsk_perception::SparseImage_<std::allocator<void> > const> const&)>, void, boost::shared_ptr<jsk_perception::SparseImage_<std::allocator<void> > const> >::invoke(boost::detail::function::function_buffer&, boost::shared_ptr<jsk_perception::SparseImage_<std::allocator<void> > const>) (
    function_obj_ptr=..., a0=...)
    at /usr/include/boost/function/function_template.hpp:153
#10 0x00007fffec310f34 in operator() (
    a0=<error reading variable: access outside bounds of object referenced via synthetic pointer>, this=0x7fffd0001048)
    at /usr/include/boost/function/function_template.hpp:1013
#11 ros::SubscriptionCallbackHelperT<boost::shared_ptr<jsk_perception::SparseImage_<std::allocator<void> > const> const&, void>::call (this=0x7fffd0001040, 
    params=...)
    at /opt/ros/groovy/include/ros/subscription_callback_helper.h:180
#12 0x00007ffff792c797 in ros::SubscriptionQueue::call() ()
   from /opt/ros/groovy/lib/libroscpp.so
#13 0x00007ffff78de5c9 in ros::CallbackQueue::callOneCB(ros::CallbackQueue::TLS*---Type <return> to continue, or q <return> to quit---
) () from /opt/ros/groovy/lib/libroscpp.so
#14 0x00007ffff78df463 in ros::CallbackQueue::callOne(ros::WallDuration) ()
   from /opt/ros/groovy/lib/libroscpp.so
#15 0x00007ffff7bbf1ea in nodelet::detail::CallbackQueue::callOne() ()
   from /opt/ros/groovy/lib/libnodeletlib.so
#16 0x00007ffff7bc0512 in nodelet::detail::CallbackQueueManager::workerThread(nodelet::detail::CallbackQueueManager::ThreadInfo*) ()
   from /opt/ros/groovy/lib/libnodeletlib.so
#17 0x00007ffff5850ce9 in thread_proxy ()
   from /usr/lib/libboost_thread.so.1.46.1
#18 0x00007ffff760ee9a in start_thread (arg=0x7fffd7fff700)
    at pthread_create.c:308
#19 0x00007ffff6e253fd in clone ()
    at ../sysdeps/unix/sysv/linux/x86_64/clone.S:112
#20 0x0000000000000000 in ?? ()

new package request: clothbag_rim_detector

I would like to add clothbag_rim_detector (Miyake, SI2013) to jsk_recognition.
I'm trying to create PR at
https://github.com/hyaguchijsk/jsk_recognition/tree/feature/clothbag_rim_detector

I've tested with Hydro/catkin and not tested with rosbuild yet,
so I will send PR after applying rosbuild.

jsk_pcl_ros nodelets subscribe tf

because of inheritance of PCLNodelet

cr_calibration fails to compile

how do I compile cr_calibration package? or is it deprecated? It seems there are several robot hardware that uses cr setup.

k-okada@kokada-t430s:~/ros/groovy/jsk-ros-pkg/jsk_recognition/cr_calibration$ make
mkdir -p bin
cd build && cmake -Wdev -DCMAKE_TOOLCHAIN_FILE=/opt/ros/groovy/share/ros/core/rosbuild/rostoolchain.cmake  ..
grep: nodes/cameracalibrator.py: そのようなファイルやディレクトリはありません
/home/k-okada/ros/groovy/jsk-ros-pkg/jsk_recognition/cr_calibration/apply_patch.sh: 4 行: [: ==: 単項演算子が予期されます
-- Configuring done
-- Generating done
CMake Warning:
  Manually-specified variables were not used by the project:

    CMAKE_TOOLCHAIN_FILE

no robot_self_filter in hydro

• official kforge is lost https://kforge.ros.org/armnavigation/armnavigation
• github port is available for groovy/rosbuild https://github.com/uos/arm_navigation/tree/master/robot_self_filter
• moveit has self_filter https://github.com/ros-planning/moveit_ros/tree/hydro-devel/perception, but it's not exported

from answers.ros.org

• http://answers.ros.org/question/69004/robot-self-model-filter-for-depth-sensor-in-ros-hydro/
• http://answers.ros.org/question/165676/take-3d-self-filter-out-of-moveit/

/selected_pointcloud と /bounding_box_marker/sected_box_array の timestamp がずれている

package://jsk_pcl_ros/launch/organized_multi_plane_segmentation.launch
を実行したときの，
/selected_pointcloud
と
/bounding_box_marker/sected_box_array
のtimestamp がずれている

renew_tracking.py fails on catkin workspace

please update load_manifest part https://github.com/jsk-ros-pkg/jsk_recognition/blob/master/jsk_pcl_ros/scripts/renew_trakcing.py#L4) as follows

PKG = 'jsk_pcl_ros'

import imp  ## for rosbuild
try:
    imp.find_module(PKG)
except:
    import roslib; roslib.load_manifest(PKG)

because renew_tracknig.py fails as follows,

-------------------------- output from roslaunch jsk_pcl_ros tracking.launch ------------------------------

INFO] [1397732968.339463573]: Loading nodelet /pcl_nodelet/particle_filter_tracker of type jsk_pcl/ParticleFilterTracking to manager /camera/camera_nodelet_manager with the following remappings:
[ INFO] [1397732968.339530273]: /pcl_nodelet/particle_filter_tracker/input -> /pcl_nodelet/voxelgrid/output
[ INFO] [1397732968.715689285]: rgb_frame_id = '/camera_rgb_optical_frame' 
[ INFO] [1397732968.715743675]: depth_frame_id = '/camera_depth_optical_frame' 
process[renew_trakcing-24]: started with pid [8204]
[ WARN] [1397732968.821515264]: Camera calibration file /home/wesleyc/.ros/camera_info/rgb_1208100007.yaml not found.
[ WARN] [1397732968.821589218]: Using default parameters for RGB camera calibration.
[ WARN] [1397732968.821667771]: Camera calibration file /home/wesleyc/.ros/camera_info/depth_1208100007.yaml not found.
[ WARN] [1397732968.821726441]: Using default parameters for IR camera calibration.

Traceback (most recent call last):
  File "/home/wesleyc/ros/hydro/src/jsk-ros-pkg/jsk_recognition/jsk_pcl_ros/scripts/renew_trakcing.py", line 4, in <module>

implement nodelet throttle/mux for PointCloud2 and Image

implement them in jsk_pcl_ros (and send the PR to pcl_ros repository).

pcl_ros has mux/demux for PointCloud2 already.

1. implement throttle for PointCloud2 on jsk_pcl_ros.
2. implement mux/demux and throttle for Image on jsk_pcl_ros.
3. send PR to pcl_ros about 1.

rewrite organized point cloud with depth_image_creator and exiting code to convert depth->pointcloud

36e5915 should be removed

2014_05_12_610_clothes.bag のサンプルをつくる

sample_610_clothes.launchをつくって，
/home/jsk/ros/data/large/2014_05_12_610_clothes.bag
をplayしてrvizで表示するものを作りましょう．

add groovy/catkin build test

BoundingBoxを表現するメッセージが必要か

3時限的なROIの表現や、認識結果などをBoundingBoxで表示するためにBoundingBoxのメッセージがあると良いか?

moveit_msgsのものを使えば良い? しかしheaderがない

$ rosmsg list | grep -i box
HLAC/BoxSizes
box_detector/Box
box_detector/BoxArray
jsk_pcl_ros_unreleased/BoxSizes
manipulation_msgs/ClusterBoundingBox
moveit_msgs/OrientedBoundingBox
$ rosmsg list | grep -i aabb
hrpsys_ros_bridge/OpenHRP_AABB
openraveros/AABB

$ rosmsg show moveit_msgs/OrientedBoundingBox
geometry_msgs/Pose pose
  geometry_msgs/Point position
    float64 x
    float64 y
    float64 z
  geometry_msgs/Quaternion orientation
    float64 x
    float64 y
    float64 z
    float64 w
geometry_msgs/Point32 extents
  float32 x
  float32 y
  float32 z

fail to generate deb on hydro due to build_depend pcl

bloom_release faild for hydro due to missing dependency for pcl
http://jenkins.jsk.imi.i.u-tokyo.ac.jp:8080/job/bloom_release/754/console
due to 21d9d09
please remove depend to pcl and there are two direction, compile optical flow component only for hydro or higher.

認識用データセットを作る

何処に置くかはチョット考え物だけど(gitに入れると重くなるか？），認識用のデータセットを作りたい

pr2で画像をbagをとるのでいいかな．１−３分ぐらいか．データを撮るのは
jsk_data/launch/pr2_record.launch でよいか．

@h-kamada エレベータホールのちょっと手前から動いていって，ボタンを（人が押して）中に入って，ボタンを見るところまで，
@aginika 部屋の中に８この布がある．机の上３，椅子の上１，床の上４，見回して一つ掴んでかごに入れるまで．
@furushchev 冷蔵庫と食洗機と洗濯機をみて回って洗濯機のボタンを押して物を入れる（最初からものは持っている），最後にチョット横に動いて箒を見る．
@mmurooka ３種類のバルブを一つづつ見る，移動してホースの先を見る，ホースは人が掴んで管に回し入れる，その時にホースの胴体が良く見えるように．

openni2_remote.launchのキャリブレーションしてない時の挙動

openni2_remote.launchのなかで，

      <node pkg="nodelet" type="nodelet" name="depth_calibration"
          args="load jsk_pcl/DepthCalibration $(arg manager) $(arg bond)"
          respawn="$(arg respawn)"
          clear_params="true"
          output="screen">
        <remap from="~input" to="/camera_remote/depth_registered/hw_registered/image_rect_raw" />
        <remap from="~camera_info" to="/camera/depth_registered/camera_info" />
        <remap from="~output" to="/camera_remote/depth_registered/hw_registered/image_rect_raw_calibrated" />
        <rosparam command="load" file="$(arg depth_calibration_file)" />
      </node>

ここのrosparamがdepth_calibration_fileに何も指定しないとエラーになってしまいます．
全個体をキャリブするというのが仕様になっている気がしますが，
そうではない個体に対してはどうしたらよいでしょうか？
ダミーのファイルを渡す？

のと，このlaunchファイル肥大化しすぎていませんか？
openni2_local.launchに対応するのがopenni2_remote.launchだという認識でしたが，
それ以上のことを単一のlaunchで行うのは正しい方向でしょうか．

後段のキャリブや画像縮小は他のファイルにして，includeする形にして，
やるかどうかはargで都度指定する，などはどうでしょうか．

Estimate OpenNI z_scale and z_offset

In order to calibrate depth value from openni sensors, we need to estimate z_scale and z_offset parameter.

z' = z_offset + z_scale * z

connection_based_nodeletのadvertiseを使ってないnodeがある。

#324 でconnection_based_nodeletが使われるようになってますが、

いくつかのコードで、advertiseのメソッドがconnection_based_nodelet
で定義されたものを使うものに更新されてない気がします。
#434 でpointcloud_screenpoint_nodelet.cppを変更してしまいましたが、

そのような変更がいくつかのコードで必要かもしれません。

Publishing same object using SelectPointCloudPublishAction twice crashes rviz

Steps to reproduce:

1. Start roscore
2. roslaunch jsk_pcl_ros tracking.launch
3. rosrun rviz rviz
4. select and object in rviz
5. click SelectPointCloudPublish Action in the SelectPointCloudPublishAction plugin.
6. re-select the same object in rviz
7. click SelectPointCloudPublish Action in the SelectPointCloudPublishAction plugin again.
   -> rviz crashes.

Reproducibility 5/5 times.
ROS version: hydro
Camera: Xtion PRO Live

Not reproducible if jsk_pcl_ros tracker is not running.

============== output from terminal running rviz ==============
wesleyc@equuleus:~/ros/hydro/src/jsk-ros-pkg/jsk_recognition/jsk_pcl_ros/include/jsk_pcl_ros$ rosrun rviz rviz
[ INFO] [1398833987.803890825]: rviz version 1.10.14
[ INFO] [1398833987.803956189]: compiled against OGRE version 1.7.4 (Cthugha)
[ERROR] [1398833987.838891948]: Skipping XML Document "/home/wesleyc/ros/hydro/src/bosch-ros-pkg/stacks/bosch_shared_autonomy/bosch_object_segmentation_gui/lib/rviz_plugin.yaml" which had no Root Element. This likely means the XML is malformed or missing.
[ INFO] [1398833988.183860078]: Stereo is NOT SUPPORTED
[ INFO] [1398833988.183949720]: OpenGl version: 4.4 (GLSL 4.4).
[ERROR] [1398833988.323733688]: Skipping XML Document "/home/wesleyc/ros/hydro/src/bosch-ros-pkg/stacks/bosch_shared_autonomy/bosch_object_segmentation_gui/lib/rviz_plugin.yaml" which had no Root Element. This likely means the XML is malformed or missing.
[ERROR] [1398833988.356534370]: Skipping XML Document "/home/wesleyc/ros/hydro/src/bosch-ros-pkg/stacks/bosch_shared_autonomy/bosch_object_segmentation_gui/lib/rviz_plugin.yaml" which had no Root Element. This likely means the XML is malformed or missing.
[ERROR] [1398833988.449641450]: Skipping XML Document "/home/wesleyc/ros/hydro/src/bosch-ros-pkg/stacks/bosch_shared_autonomy/bosch_object_segmentation_gui/lib/rviz_plugin.yaml" which had no Root Element. This likely means the XML is malformed or missing.
[ERROR] [1398833989.072018867]: Skipping XML Document "/home/wesleyc/ros/hydro/src/bosch-ros-pkg/stacks/bosch_shared_autonomy/bosch_object_segmentation_gui/lib/rviz_plugin.yaml" which had no Root Element. This likely means the XML is malformed or missing.
[ERROR] [1398833989.595340562]: Skipping XML Document "/home/wesleyc/ros/hydro/src/bosch-ros-pkg/stacks/bosch_shared_autonomy/bosch_object_segmentation_gui/lib/rviz_plugin.yaml" which had no Root Element. This likely means the XML is malformed or missing.
[ INFO] [1398834041.818093321]: num > 493!
[ INFO] [1398834043.394398092]: num > 493!
[ INFO] [1398834044.216829427]: num > 493!
[ INFO] [1398834048.470367020]: num > 406!
Segmentation fault (core dumped)

================== output from terminal running tracker.launch ===============
[ INFO] [1398834270.537679670]: Tracking.. 46312
[ INFO] [1398834270.579608361]: Tracking.. 46206
[ INFO] [1398834270.629592114]: Tracking.. 46334
[ INFO] [1398834270.668419490]: Tracking.. 46163
[ INFO] [1398834270.716657668]: Tracking.. 46382
[ INFO] [1398834270.764416083]: Tracking.. 46136
[ INFO] [1398834270.807682917]: Tracking.. 46280
[ INFO] [1398834270.851970664]: Tracking.. 46319
[ INFO] [1398834270.891480243]: Tracking.. 45983
[ INFO] [1398834270.942302903]: Tracking.. 46107
[ INFO] [1398834270.992710290]: Tracking.. 46175
[ INFO] [1398834271.041415621]: Tracking.. 45818
[ INFO] [1398834271.090642854]: Tracking.. 46124
[ INFO] [1398834271.176652822]: Tracking.. 46168

image_rotate supporting nodelet and tf2

First, I will send PR to ros_perception.
and after that, I will add that to jsk_pcl_ros as a compensation until the PR will be committed.

motivation:
image_rotate inside of PR2 is pretty heavy because of tf and nodelet communication(?).

jsk_pcl_ros documentation

documentを作る。

• 何をするためのコードか
• パラメータの一覧
• ユースケース
• デモ画像・動画

libjsk_pcl_ros.so: error: undefined reference to 'cv_bridge::CvImage::toImageMsg() const'

I got the error as below in hydro.
What should I do?

When I added the run_depend and build_depend to the cv_bridge and add cv_bridge to catkin.cmake, the make worked.But Travis seems to pass it, doesn't it?

[100%] Building CXX object jsk-ros-pkg/jsk_visualization/jsk_interactive_markers/jsk_interactive_marker/CMakeFiles/door_foot.dir/src/interactive_marker_helpers.cpp.o
Linking CXX executable /home/******/ros/hydro/devel/lib/jsk_interactive_marker/marker_6dof
/home/******/ros/hydro/devel/lib/libjsk_pcl_ros.so: error: undefined reference to 'cv_bridge::toCvShare(boost::shared_ptr<sensor_msgs::Image_<std::allocator<void> > const> const&, std::basic_string<char, std::char_traits<char>, std::allocator<char> > const&)'
/home/******/ros/hydro/devel/lib/libjsk_pcl_ros.so: error: undefined reference to 'cv_bridge::CvImage::toImageMsg() const'
collect2: ld はステータス 1 で終了しました
make[2]: *** [/home/******/ros/hydro/devel/lib/jsk_interactive_marker/marker_6dof] エラー 1
make[1]: *** [jsk-ros-pkg/jsk_visualization/jsk_interactive_markers/jsk_interactive_marker/CMakeFiles/marker_6dof.dir/all] エラー 2
make[1]: *** 未完了のジョブを待っています....
[100%] Building CXX object jsk-ros-pkg/jsk_visualization/jsk_interactive_markers/jsk_interactive_marker/CMakeFiles/interactive_marker_interface.dir/src/interactive_marker_utils.cpp.o
/home/******/ros/hydro/src/jsk-ros-pkg/jsk_visualization/jsk_interactive_markers/jsk_interactive_marker/src/interactive_marker_helpers.cpp: 関数 ‘visualization_msgs::InteractiveMarker im_helpers::makePosedMultiMeshMarker(const char*, const PoseStamped&, const std::vector<geometry_msgs::PoseStamped_<std::allocator<void> > >&, const std::vector<std::basic_string<char> >&, const float&, bool)’ 内:
/home/******/ros/hydro/src/jsk-ros-pkg/jsk_visualization/jsk_interactive_markers/jsk_interactive_marker/src/interactive_marker_helpers.cpp:703:160: 警告: enumeral mismatch in conditional expression: ‘visualization_msgs::InteractiveMarkerControl_<std::allocator<void> >::<anonymous enum>’ vs ‘visualization_msgs::InteractiveMarkerControl_<std::allocator<void> >::<anonymous enum>’ [デフォルトで有効]
Linking CXX executable /home/******/ros/hydro/devel/lib/jsk_interactive_marker/door_foot
/home/******/ros/hydro/devel/lib/libjsk_pcl_ros.so: error: undefined reference to 'cv_bridge::toCvShare(boost::shared_ptr<sensor_msgs::Image_<std::allocator<void> > const> const&, std::basic_string<char, std::char_traits<char>, std::allocator<char> > const&)'
/home/******/ros/hydro/devel/lib/libjsk_pcl_ros.so: error: undefined reference to 'cv_bridge::CvImage::toImageMsg() const'
collect2: ld はステータス 1 で終了しました
make[2]: *** [/home/******/ros/hydro/devel/lib/jsk_interactive_marker/door_foot] エラー 1
make[1]: *** [jsk-ros-pkg/jsk_visualization/jsk_interactive_markers/jsk_interactive_marker/CMakeFiles/door_foot.dir/all] エラー 2

color_histogram_matcher_nodeletでの類似度の可視化

可視化のために、color_histogramの類似度の可視化をしてみようと思います。
何パターンか実装したので、画像を載せます

Cannot start multiple particle tracker nodes to track multiple objects

Starting particle tracking node under a namespace does not work, so cannot start multiple tracking nodes in separtated namespaces to track multiple objects.

release 0.1.4

ready?

travisでjsk_pcl_rosがメモリ不足で死にすぎる

並列数2でもメモリ不足で死ぬのはさすがに問題があるので問題を探りたい。
最適化のオプションをなくしてみる?

Particle fitler tracker does not use color information

particle filter tracker does not use color information so it often starts tracking the arm when the object is picked up, even when the arm and the object have very different colors.