0.1.22 (2014-09-24)

• Disable ssl when calling git
• Contributors: Ryohei Ueda

0.1.21 (2014-09-20)

• Add more diagnostics to OrganizedMultiPlaneSegmentation and fix global hook for ConvexHull
• Contributors: Ryohei Ueda

0.1.20 (2014-09-17)

0.1.19 (2014-09-15)

0.1.18 (2014-09-13)

• add git to build_depend of jsk_libfreenect2
• Contributors: Ryohei Ueda

0.1.17 (2014-09-07)

• add mk/git to build_depend
• Contributors: Kei Okada

0.1.16 (2014-09-04)

• do not use rosrun in the script of jsk_perception/src/eusmodel_template_gen.sh
• Contributors: Ryohei Ueda

0.1.14 (2014-08-01)

0.1.13 (2014-07-29)

0.1.12 (2014-07-24)

• fix to use catkin to link rospack
• Contributors: Dave Coleman, Kei Okada

0.1.11 (2014-07-08)

• jsk_perception does not depends on pcl, but depends on eigen and tf
• Contributors: Ryohei Ueda

0.1.10 (2014-07-07)

• adding oriented_gradient_node
• add calc_flow program to calc optical flow
• Contributors: Hiroaki Yaguchi, Ryohei Ueda

0.1.9 (2014-07-01)

0.1.8 (2014-06-29)

• initialize _img_ptr at first
• convert color image to GRAY
• add nodelet to detect circles based on hough transformation
• add program to compute color histogram (rgb and hsv color space)
• maked configure_file to create imagesurf, imagestar and imagebrisk automatically
• added the programs to use cv_detection
• Contributors: Ryohei Ueda, Yu Ohara, Yusuke Furuta

0.1.7 (2014-05-31)

0.1.6 (2014-05-30)

0.1.5 (2014-05-29)

• add service interface with sensor_msgs/SetCameraInfo to camshiftdemo, not only mouse selection.
• Contributors: Ryohei Ueda

0.1.4 (2014-04-25)

• add sparse_image program to jsk_percepton
• make edge_detector nodelet class
• Contributors: Ryohei Ueda, furushchev
• Merge pull request #47 from k-okada/add_rosbuild
• Contributors: Kei Okada

0.1.3 (2014-04-12)

0.1.2 (2014-04-11)

0.1.1 (2014-04-10)

• catkinize jsk_perception
• check initialization in check_subscribers function
• change callback function names for avoiding the same name functions
• add edge_detector.launch
• change debug message
• rename type -> atype
• fix minor bug
• change for treating multiple objects in one ObjectDetection.msg
• add test programs
• add rosbuild_link_boost for compile on fuerte/12.04 , see Issue #224, thanks tnakaoka
• add rectangle_detector, based on http://opencv-code.com/tutorials/automatic-perspective-correction-for-quadrilateral-objects/
• update hoguh_lines
• use blur before canny
• add image_proc modules from opencv samples
• change error_threshold max 200 -> 2000
• add :detection-topic keyword to (check-detection)
• replace sleep to :ros-wait for making interruptible
• add scripts for speaking english
• speak before sleep
• add to spek we're looking for...
• print out debug info
• turtlebot/ros pdf
• add ros/turtlebot-logo images #173
• update japanese speaking
• modify parameter definition. parameter should not be overwritten.
• add option publish-objectdetection-marker
• add slot :diff-rotation in detection_interface.l
• do not create ros::roseus object by load detection_interface.l
• publish tf from sensor frame to detected object pose
• update objectdetection-marker program for new detection_interface
• publish tf and markers, add messages
• print out error value
• fix segfault
• suppor rpy style in relative_pose, status:closed #139
• add :target-object keyword to check-detection
• fix : project3dToPixel was removed in groovy
• update to use cv_bridge
• fix for groovy, use cv_bridge not CvBridge
• fix: speak content
• fix: speak-jp
• fix template location
• add microwave detection sample
• add speak-name for speaking japanease object name
• add speak words
• update detction_interface.l for single detection and speak flag
• add solve-tf parameter for not using tf
• add frame_id for coordinates
• add detection_interface.l for using point_pose_extractor
• remove euclidean_cluster,plane_detector and color_extractor from jsk_perception, they are supported in tabletop and pcl apps should go into jsk_pcl_ros
• add max_output
• add opencv2 to rosdep.yaml for compatibility
• update to fit opencv2 electric/fuerte convention
• fix for fuerte see https://code.ros.org/trac/ros/ticket/3955
• add size check
• fix btVector3 -> tf::Vector3
• fix remove define KdTreePtr
• fix style: support ROSPACK_API_V2 (fuerte)
• support ROSPACK_API_V2 (fuerte)
• fix for pcl > 1.3.0, pcl::KdTree -> pcl::search::KdTree, pcl::KdTreeFLANN -> pcl::search::KdTree
• remove explicit dependency to eigen from jsk_perception
• add whilte_balance_param.yaml
• add publish_array for publishing pointsarray
• move posedetectiondb/SetTemplate -> jsk_perception/SetTemplate
• add color_extractor, plane_detector, euclidean_clustering for jsk_perception
• fixed the package name of WhiteBalance.srv
• add eigen to dependency
• add white_balance_converter to jsk_perception
• change msg from face_detector_mono/Rect -> jsk_perception/Rect. I couldn't find set_serch_rect string under jsk-ros-pkg
• node moved from virtual_camera
• check if the matched region does not too big or too small
• add dynamic reconfigure for point_pose_extractor
• split launch for elevator_navigation, to test modules
• fix for oneiric
• fix for users who does not have roseus in their PATH
• ns can't be empty string in launch xml syntax
• commit updates for demo
• added tv-controller with ut logo
• added tv-controller with ut logo
• fixed the size of wrap image, which is calcurated from input (width/height)
• add to write wrapped image
• add error handling and output template file
• add opencv-logo2.png
• add lipton milktea model, auto generated file prefix .launch -> .xml to avoid listed by auto complete
• add sharp rimokon with ist logo
• changed variable name client -> clients
• add sharp tv controller to sample
• add sample for detection launcher generator
• use try to catch assertions
• set Zero as distortionMatrix, because ImageFeature0D.image is rectified
• fixed the box pose in debug image
• changed code for generate SIFT template info
• use projectionMatrix instead of intrinsicMatrix in solvePnP, remove CvBridge -> cv_bridge
• fix to work without roseus path in PATH
• fix relative pose, object coords to texture coords
• update generation script of SIFT pose estimation launcher, relative pose is not correct
• update eusmodel->sift_perception script
• change detection launch generation script to use jsk_perception/point_pose_extractor
• add std namespace appropriately
• update initialize template method
• publish the debug_image of point_pose_extractor
• chnage the output frame id when using only one template
• change threashold for detectiong object
• use /ObjectDetection_agg instead of /ObjectDetection
• add _agg output topic for debug and logging
• add debug message, set lifetime to 1 sec
• add objectdetection-marker.l
• add relative pose parameter to point_pose_extractor.cpp
• change the PutText region
• update sample launch file, point pose extractor do not subscribe input topics when output is not subscribed
• add viewer_window option to disable the OpenCV window
• empty window name to disable window, point_pose_extractor
• move posedetectiondb to jsk_visioncommon
• moved jsk_vision to jsk_visioncommon
• Contributors: HiroyukiMikita, Kei Okada, Haseru Chen, Yuki Furuta, Yuto Inagaki, Kazuto Murai, Manabu Saito, Rosen Dinakov, Youhei Kakiuchi