 matlabbe
|
0f03b69144
|
ros: Fixed detection not working if gui was false (#100)
|
2020-01-08 17:42:19 -05:00 |
|
|
matlabbe
|
aefb42dc57
|
Fixed wrong detector index chosen when reading ini if index is over 9.
|
2020-01-08 17:41:24 -05:00 |
|
|
matlabbe
|
5489d5f4c4
|
removed debugging logs
|
2019-09-02 20:03:19 -04:00 |
|
|
matlabbe
|
d661148694
|
fixed https://github.com/introlab/rtabmap_ros/issues/337
|
2019-09-02 20:00:54 -04:00 |
|
|
matlabbe
|
bde868d1cc
|
Added PnP approach (now default) to compute 3D transform (requiring only single valid depth point instead of 3). Added arguments to launch files for convenience. Removed fake map/odom frames from find_object_3d.launch (so that it can be more easily included in other launch files).
|
2019-08-10 15:31:12 -04:00 |
|
|
matlabbe
|
f43bc943f8
|
Refactored how depth is used to compute 3D transform (making sure the same depth image is used with same stamp and frame_id). Fixed CameraTcpServer ingoring incoming connections with Qt5. Added approx_sync parameter to ros.
|
2019-08-09 21:45:27 -04:00 |
|
|
matlabbe
|
413cb6820d
|
Fixed build with OpenCV4
|
2019-06-17 11:43:26 -04:00 |
|
|
matlabbe
|
5047163c09
|
catkin: updated dependencies for message generation
|
2019-03-31 17:42:05 -04:00 |
|
|
matlabbe
|
ef0c565b7e
|
added 3d example with zed, added bgra support on rgb/depth callback
|
2019-01-28 16:37:35 -05:00 |
|
|
matlabbe
|
21c80c9d82
|
Merge pull request #74 from nuclearsandwich/patch-1
Fix markdown syntax in OpenCV link.
|
2018-12-18 22:01:30 +00:00 |
|
|
matlabbe
|
7a33accddd
|
Added new msg DetectionInfo (#71)
|
2018-11-21 14:11:28 -05:00 |
|
|
matlabbe
|
53a6f933d1
|
Update .appveyor.yml
|
2018-10-23 15:03:45 -04:00 |
|
|
matlabbe
|
ddc86d1bef
|
Update README.md
|
2018-10-24 05:58:11 +12:00 |
|
|
matlabbe
|
b4d9bfc468
|
CameraROS: making image conversion more general to support more formats than bgr8 and rgb8 (#69).
|
2018-10-10 17:01:33 -04:00 |
|
|
matlabbe
|
784691a070
|
fixed compilation error with HAVE_OPENCV_CUDAFEATURES2D
|
2018-09-18 20:14:15 -04:00 |
|
|
matlabbe
|
556bf5f4cc
|
Session: Compression of the vocabulary to avoid QByteArray max 2GB limitation
|
2018-09-10 18:57:51 -04:00 |
|
|
matlabbe
|
c6cf667645
|
fixed feature detector/descriptor not updated after loading a session
|
2018-09-06 17:51:11 -04:00 |
|
|
matlabbe
|
eefd6c6ff6
|
GU mode: use default config file if argument --config and --session are not set
|
2018-08-31 12:01:03 -04:00 |
|
|
matlabbe
|
0424bf181c
|
If a config file is set on arguments (or if there are custom parameters), parameters overwrite those in a loaded session.
|
2018-08-31 11:53:34 -04:00 |
|
|
matlabbe
|
ade5dd6729
|
Save/Load vocabulary options can read/write in session binary format
|
2018-08-30 12:43:16 -04:00 |
|
|
matlabbe
|
9d0c03f661
|
Fixed saving session without objects (to save fixed vocabulary in binary format)
|
2018-08-30 12:21:51 -04:00 |
|
|
matlabbe
|
147a8a0dac
|
refactored tf generation
|
2018-07-30 16:10:09 -04:00 |
|
|
matlabbe
|
cb349e557f
|
Suppressed some cmake policy warnings
|
2018-07-30 15:54:01 -04:00 |
|
|
matlabbe
|
6fa5e0c8f4
|
fixed -ltcmalloc not defined error
|
2018-07-30 15:31:49 -04:00 |
|
|
matlabbe
|
9943c28b88
|
Fixed build when tcmalloc is not found
|
2018-07-30 15:27:03 -04:00 |
|
|
matlabbe
|
98e605e3d1
|
Merge branch 'nhudinh2103-fix-memory-leak-linux'
|
2018-07-30 14:42:47 -04:00 |
|
|
matlabbe
|
1abda367d6
|
fixed some memory leaks (#55)
|
2018-07-04 12:28:58 -04:00 |
|
|
matlabbe
|
2cb9120f74
|
ROS: fixed TF when multiDetection is enabled (#54)
|
2018-05-29 14:53:05 -04:00 |
|
|
matlabbe
|
890ad0fe0a
|
updated appveyor
|
2018-05-28 15:19:08 -04:00 |
|
|
matlabbe
|
b55ff1025c
|
updated appveyor
|
2018-05-28 15:14:18 -04:00 |
|
|
matlabbe
|
fc8ea483ce
|
added appveyor config
|
2018-05-28 15:11:34 -04:00 |
|
|
matlabbe
|
3e1f4e6669
|
Packaging: fixed qt5 plugins copy
|
2018-05-28 14:56:10 -04:00 |
|
|
matlabbe
|
ab1518b167
|
Fixing #52
|
2018-05-22 11:16:39 -04:00 |
|
|
matlabbe
|
cdcb8de9e7
|
increased version to 0.6.2
|
2017-11-07 11:07:35 -05:00 |
|
|
matlabbe
|
a0e0e763a5
|
Added example drawing object rectangles on image (#48)
|
2017-10-10 19:03:03 -04:00 |
|
|
matlabbe
|
0ada2cc8cb
|
fixed #47
|
2017-10-09 19:11:41 -04:00 |
|
|
matlabbe
|
d99426eb85
|
fixed #46
|
2017-10-04 10:58:00 -04:00 |
|
|
matlabbe
|
ca01bf55d4
|
OpenCV3: added error msg with bruteforce GPU if not built with cudafeature2d. Fixed compilation error with OpenCV3+cudafeature2d module.
|
2017-09-14 21:41:00 -04:00 |
|
|
matlabbe
|
070bd474d6
|
Update README.md
|
2017-08-07 18:59:47 -04:00 |
|
|
matlabbe
|
ca1e1dccde
|
Homography: added maxIterations and configences parameters (OpenCV3). Updated doubleSpinBox decimals/singleStep.
|
2017-04-27 14:40:04 -04:00 |
|
|
matlabbe
|
419a8a6f51
|
Fixed crash when setting max features (#34)
|
2017-03-31 13:29:54 -04:00 |
|
|
matlabbe
|
6337d4efb6
|
Added General/debug parameter
|
2017-03-27 13:56:53 -04:00 |
|
|
matlabbe
|
f722087bf5
|
Update README.md
|
2017-03-24 10:25:17 -04:00 |
|
|
matlabbe
|
b5565bb539
|
Update README.md
|
2017-03-24 10:24:51 -04:00 |
|
|
matlabbe
|
d68a755c7a
|
Update README.md
|
2017-03-24 10:24:40 -04:00 |
|
|
matlabbe
|
dc042fe418
|
Added General/autoScreenshotPath option
|
2017-03-15 11:48:36 -04:00 |
|
|
matlabbe
|
1f44524b9d
|
Fixed crash when selecting hamming flann distance, then change to bruteforce NN, then change to float descriptor
|
2016-09-22 12:24:25 -04:00 |
|
|
matlabbe
|
77b95d6b8c
|
Update Vocabulary.cpp
|
2016-09-03 19:24:54 -04:00 |
|
|
matlabbe
|
c899c2033d
|
catkin: added dependency to generated msgs for the lib
|
2016-06-13 12:02:06 -04:00 |
|
|
matlabbe
|
48e6ea0147
|
Replaced obj_frame_prefix by object_prefix
|
2016-06-13 11:21:36 -04:00 |
|
|
matlabbe
|
881f70377f
|
Added CATKIN_ENABLE_TESTING to detect catkin on dev build on ROS buildfarm
|
2016-05-09 13:45:49 -04:00 |
|
|
matlabbe
|
a5c25df708
|
Added CATKIN_SKIP_TESTING to detect catkin on dev build on ROS buildfarm
|
2016-05-09 12:34:41 -04:00 |
|
|
matlabbe
|
dcd811e3da
|
Depends on qt_gui_cpp instead of qt_gui_core metapackage
|
2016-05-09 11:15:32 -04:00 |
|
|
matlabbe
|
be29572ed1
|
Fixed catkin detection by cmake for binary release on Kinetic
|
2016-05-08 16:34:41 -04:00 |
|
|
matlabbe
|
e187dae127
|
package.xml: added qt_gui_core for Qt dependency
|
2016-05-06 18:31:43 -04:00 |
|
|
matlabbe
|
52ce5076db
|
Auto find qt version (qt5 has priority)
|
2016-05-06 16:54:18 -04:00 |
|
|
matlabbe
|
5c21a76653
|
travis: disabled email notifications
|
2016-04-28 10:11:04 -04:00 |
|
|
matlabbe
|
662ff22c1d
|
Update README.md
|
2016-04-28 09:08:30 -04:00 |
|
|
matlabbe
|
f5ca03f0a6
|
travis dist: trsuty
|
2016-04-28 09:02:30 -04:00 |
|
|
matlabbe
|
1f2eb81fa0
|
modified .travis.yml script
|
2016-04-28 08:55:31 -04:00 |
|
|
matlabbe
|
2601a7dc54
|
travis cmake
|
2016-04-28 08:46:26 -04:00 |
|
|
matlabbe
|
4760802f1d
|
Added .travis.yml
|
2016-04-28 08:41:49 -04:00 |
|
|
matlabbe
|
b3056ca3b7
|
Fixed build for 2.4.12 (miniflann interface changed)
|
2016-03-06 14:12:28 -05:00 |
|
|
matlabbe
|
467b82a120
|
Update find_object_3d_kinect2.launch
|
2016-02-22 19:08:22 -05:00 |
|
|
matlabbe
|
4b68201a23
|
ROS: added kinect2 example
|
2016-02-22 18:50:37 -05:00 |
|
|
matlabbe
|
91bce9ac69
|
Added "NearestNeighbor/7ConvertBinToFloat" parameter: This will convert binary descriptors to float so that vocabulary approaches with KD-Tree or K-means can be used with binary descriptors like Brief/ORB/FREAK (use MANHATTAN_L1 too)
|
2016-02-01 12:53:16 -05:00 |
|
|
matlabbe
|
138d004169
|
Added RootSIFT (issue #8): see "Feature2D/SIFT_rootSIFT" parameter
|
2016-01-19 15:02:17 -05:00 |
|
|
matlabbe
|
9525a55617
|
fixed empty image fatal error when updating objects on ROS
|
2016-01-19 00:03:10 -05:00 |
|
|
matlabbe
|
3476d91bb4
|
clear in/out label when updating objects
|
2016-01-17 18:02:40 -05:00 |
|
|
matlabbe
|
8b711782fd
|
Qt5: fixed crash in PdfPlotCurve::setData()
|
2016-01-17 17:33:07 -05:00 |
|
|
matlabbe
|
55b47aa95d
|
fixed build error (<QtGui/QApplication> not found) with Qt5
|
2016-01-17 16:33:30 -05:00 |
|
|
matlabbe
|
2c0e121d82
|
Fixed crash (OpenCV Error: Unsupported format or combination of formats) when both detector and descriptors are changed AND new type is not compatible with nearest-beighbor approach used (float vs binary)
|
2016-01-17 15:21:03 -05:00 |
|
|
matlabbe
|
706f83ffd4
|
Added ASSERT before detecting features to make sure that an image exist and is CV_8UC1 type
|
2016-01-17 14:16:19 -05:00 |
|
|
matlabbe
|
52387f27b3
|
Fixed double re-detection when the camera is not running and objects were updated
|
2016-01-17 14:01:36 -05:00 |
|
|
matlabbe
|
9cc9868820
|
Fixed double features re-extraction when both detector and descriptor types are changed at the same time.
|
2016-01-17 13:50:59 -05:00 |
|
|
matlabbe
|
ccb679e7c2
|
OpenCV3: added RHO homography type "Homography/method"
|
2016-01-17 12:24:42 -05:00 |
|
|
matlabbe
|
c6ee7a27ef
|
Fixed crash when computing homographies (std::vector() out of bound error)
|
2016-01-17 11:26:12 -05:00 |
|
|
matlabbe
|
6dc66d5730
|
Fixed doubleSpinBox max value when default value=0
|
2015-12-18 17:42:38 -05:00 |
|
|
matlabbe
|
0a22ed804c
|
Version 0.6.1
|
2015-11-30 01:15:58 -05:00 |
|
|
matlabbe
|
cff1d4eac7
|
Added "--tcp_threads" argument for multi-threaded detections (on multiple ports)
|
2015-11-30 00:27:57 -05:00 |
|
|
matlabbe
|
164da72169
|
Added Qt5 support ("cmake -DFINDOBJECT_QT_VERSION=5 ..")
|
2015-11-29 18:39:20 -05:00 |
|
|
matlabbe
|
ae7b79629b
|
Removed not used pcl_ros dependency
|
2015-11-12 15:25:37 -05:00 |
|
|
matlabbe
|
bcc74f4a80
|
issue #5: select default detector/descriptor if a not available one is set in config
|
2015-08-07 23:29:52 -04:00 |
|
|
matlabbe
|
730a0bad74
|
Update README.md
|
2015-08-06 09:40:54 -04:00 |
|
|
Mathieu Labbé
|
028bfcf52e
|
Updated version to 0.6.0
|
2015-07-10 16:00:34 -04:00 |
|
|
Mathieu Labbé
|
181b00bf22
|
fixed crash when selecting SIFT detector while DAISY descriptor is used.
|
2015-07-10 15:47:28 -04:00 |
|
|
Mathieu Labbé
|
4b5da36512
|
Fixed build with OpenCV 3 CUDA (issue #3)
|
2015-07-10 14:27:11 -04:00 |
|
|
matlabbe
|
dd14e5e1ca
|
Added Load/Save vocabulary actions in File menu (in console mode, see "--vocabulary" option) (issue #2)
|
2015-07-07 16:49:38 -04:00 |
|
|
matlabbe
|
a4a9b8ca5f
|
Removed IplImage when using OpenCV3 (issue #3)
|
2015-07-07 11:09:03 -04:00 |
|
|
Mathieu Labbé
|
bb097b2bab
|
fixed truncation double->float warnings
|
2015-07-06 14:22:48 -04:00 |
|
|
matlabbe
|
f1bba5b8fb
|
Added wordID in shown keypoint info, improved fixed vocabulary behavior
|
2015-06-24 18:07:52 -04:00 |
|
|
matlabbe
|
b4e97e450d
|
First draft of a fixed vocabulary (issue #2)
|
2015-06-23 18:21:30 -04:00 |
|
|
matlabbe
|
b07239a000
|
Added OpenCV 3 support (CUDA module not yet tested)
|
2015-06-23 16:38:47 -04:00 |
|
|
matlabbe
|
a50929b0fc
|
Added a decimal to double parameters
|
2015-06-10 19:11:37 -04:00 |
|
|
matlabbe
|
388236e5c4
|
Added "--images_not_saved" argument to not keep images in RAM after the features are extracted (only in console mode)
|
2015-06-09 21:41:34 -04:00 |
|
|
matlabbe
|
9754130ec8
|
ignore ".." and "." subdirectories
|
2015-06-04 09:41:57 -04:00 |
|
|
matlabbe
|
76b6ff4b7e
|
Added option to recursively parse subdirectories when loading objects (File->"Load objects...")
|
2015-06-03 21:46:35 -04:00 |
|
|
matlabbe
|
34f0262bfd
|
Update README.md
|
2015-05-20 17:09:45 -04:00 |
|
|
matlabbe
|
f894b758d7
|
Update README.md
|
2015-05-20 16:59:23 -04:00 |
|
|
matlabbe
|
c05b208cde
|
Added auto detection of catkin
|
2015-05-20 16:55:56 -04:00 |
|