代码拉取完成,页面将自动刷新
--write_pose
renamed as --write_keypoint
and it also applies to face and/or hands.--write_pose_json
renamed as --write_keypoint_json
and it also applies to face and/or hands.--write_pose_format
renamed as --write_keypoint_format
and it also applies to face and/or hands.Too many resources requested for launch
.Maximum
instead of Nms
, since there is only 1 person / detection.Datum::scaleRatios
to save the relative scale ratio when multi-scale).--camera_fps
so generated video (--write_video
) will use that frame rate.numberParts
for Reshape
, they deduce the value.--write_keypoint_json
) are now saved as pose_keypoints
, face_keypoints
, hand_left_keypoints
, and hand_right_keypoints
. They all were previously saved as body_parts
.--num_scales
renamed as --scale_number
.--hand_
and --face_
respectively.--process_real_time
threw error with webcam.--body_parts
. Now they are called --face_keypoints
.openpose/utilities/macros.hpp
moved to openpose/utilities/macros.hpp
.OpenPose
name.Tutorial/Wrapper
VS linking errors.examples/tutorial_wrapper/
.-Wsign-compare
and removed in code.USE_CUDNN
is no longer required; USE_CAFFE
and USE_CUDA
(replacing the old CPU_ONLY
) are no longer required to use the library, only to build it. In addition, Boost, Caffe, and its dependencies have been removed from the OpenPose header files. Only OpenCV include and lib directories are required when building a project using OpenPose.USE_CAFFE
and/or USE_CUDA
are not enabled, although it will give an error saying they are required.tutorial_wrapper/
examples.-1
option for --net_resolution
in order to auto-select the best possible aspect ratio given the user input.--disable_multi_thread
flag in order to allow debug and/or highly reduce the latency (e.g., when using webcam in real-time).--resolution
renamed as --output_resolution
.core/
, pose/
, face/
, and hand/
modules.CPU_ONLY
changed by USE_CUDA
to keep format.--heatmaps_scale
to OpenPoseDemo, added option not to scale the heatmaps, and added custom float
format to save heatmaps in floating format.--write_json
as new version of --write_keypoint_json
. It includes the body part candidates (if enabled), as well as any extra information added in the future (e.g., person ID).--write_json
.PoseParameters
splitted into PoseParameters
and PoseParametersRender
and const parameters turned into functions for more clarity.getAverageScore
and getKeypointsArea
.--write_json
uses less hard disk space (enters and tabs removed).PROFILER_ENABLED
option (CMake/Makefile.config) and --profile_speed
flag._bin
by .bin
.number_people_max 1
to avoid errors (as it assumes only 1 person per image).windows/
version. CMake is the only Windows version available.--write_json
includes 3-D keypoints.--3d_views
added to allow --image_dir
and --video
allow loading stereo images.--camera_resolution
applicable to --flir_camera
.--write_video
) compatible with multi-camera setting. It will save all the different views concatenated.--3d_min_views
to select minimum number of cameras required for 3-D reconstruction.n
times faster for n
cameras (multi-threaded). If the number of cameras is greater than the number of the computer threads, the speed up might not be exactly n
times.no_display
renamed as display
, able to select between NoDisplay
, Display2D
, Display3D
, and DisplayAll
.*_keypoints
by *_keypoints_2d
to avoid confusion with 3d ones in --write_json
output file.net/
module in order to reduce core/
number of classes and files and for future scalability.COCO
model.flir_camera_index
to allow running on all the cameras at once, or only on 1 camera at the time.frame_keep_distortion
not to undistort the images. E.g., useful when recording images for camera calibration.Release
mode by default.checkEQ
from tight loop in bodyPartConnectorBase, which took a huge time exponential to the number of people.scaleKeypoints(Array<float>& keypoints, const float scale)
also accepts 3D keypoints.PoseExtractor
class to contain future ID and tracking algorithms as well as the current OpenPose keypoint detection algorithm.tracking
and identification
modules (for now disabled but available in the source code), including PersonIdExtractor
and PersonTracker
. PersonIdExtractor
includes greedy matrix OP-LK matching.--write_keypoint_json
removed (--write_json
is the equivalent since version 1.2.1).WITH_OPENCV_WITH_OPENGL
flag to tell whether to use OpenGL support for OpenCV.net_resolution
for image_dir
in CPU/OpenCL versions.--output_resolution
).scaleKeypoints
, other than scaleKeypoints(Array<float>& keypoints, const float scale)
, renamed as scaleKeypoints2d
.(W)PoseExtractor
renamed to (W)PoseExtractorNet
to distinguish from new PoseExtractor
. Analogously with (W)FaceExtractorNet
and (W)HandExtractorNet
.tracking
directory moved to main openpose directory.--output_resolution
is not the default one.--part_candidates
) are saved with the same scale than the final keypoints itself.--number_people_max
) that provoked core dumped if lots of values equal to the threshold.--tracking
flag).BUILD_UNITY_SUPPORT
on CMake, which enables special Unity code so it can be built as a Unity plugin.flags.hpp
that includes all OpenPose flags, removing the need to copy them repeatedly on each OpenPose example file.tutorial_wrapper
as tutorial_api_cpp
as well as new examples were added.tutorial_python
as tutorial_api_python
as well as new examples were added.tutorial_thread
as tutorial_api_thread
, focused in the multi-thread mechanism.tutorial_pose
, the directory tutorial_api_cpp
includes much cleaner and commented examples.--no_display
flag for the examples that does not use OpenPose output.--frame_step
to allow the user to select the step or gap between processed frames. E.g., --frame_step 5
would read and process frames 0, 5, 10, etc.COCO_CHALLENGE
variable turned into user configurable flag --maximize_positives
.--verbose
to plot the progress.--fps_max
to limit the maximum processing frame rate of OpenPose (useful to display results at a maximum desired speed).--prototxt_path
and --caffemodel_path
to allow custom ProtoTxt and CaffeModel paths.--face_detector
and --hand_detector
, that enable the user to select the face/hand rectangle detector that is used for the later face/hand keypoint detection. It includes OpenCV (for face), and also allows the user to provide its own input. Flag --hand_tracking
is removed and integrated into this flag too.
y checks to avoid --frame_last
to be smaller than --frame_first
or higher than the number of total frames.--upsampling_ratio
, which controls the upsampling than OpenPose will perform to the frame before the greedy association parsing algorithm.--body
(replacing --body_disable
), which adds the possibility of disabling the OpenPose pose network but still running the greedy association parsing algorithm (on top of the user heatmaps, see the associated tutorial_api_cpp
example).utilities/keypoints.hpp
.resizeFixedAspectRatio
can take already allocated memory (e.g., faster if target is an Array object, no intermediate cv::Mat required).utilities/keypoints.hpp
and Array to simplify keypoint post-processing.#!/bin/bash
to tell the terminal that they are bash scripts.--write_video
) can be generated from images (--image_dir
), as long as they maintain the same resolution.--write_video_3d
flag.write_video_with_audio
to enable saving these output MP4 videos with audio.CameraMatrixInitial
field into the XML calibration files to keep the information of the original camera extrinsic parameters when bundle adjustment is run.poseNetOutput
to Datum so that user can introduce his custom network output.errorWorker()
instead of error()
) lead to stopping the threads and reporting the error from the main thread, while exceptions in destructors (errorDestructor()
instead of error()
) are reported with std::cerr but not thrown as std::exceptions.tutorial_developer/python_2_pose_from_heatmaps.py
was using 2 scales starting at -1x736, changed to 1 scale at -1x368.Forward
to all net classes that automatically selects between CUDA, OpenCL, or CPU-only version depending on the defines.--camera_fps
flag by --write_video_fps
, given that it was a confusing name: It did not affect the webcam FPS, but only the FPS of the output video. In addition, default value changed from 30 to -1.--hand_tracking
is a subcase of --hand_detector
, so it has been removed and incorporated as --hand_detector 3
.--frame_keep_distortion
as --frame_undistort
, which performs the opposite operation (the default value has been also changed to the opposite).--camera_parameter_folder
as --camera_parameter_path
because it could also take a whole XML file path rather than its parent directory.--scale_gap
changed from 0.3 to 0.25.scripts/
directory. Only models/getModels.sh and the *.bat
files are kept under models/
and 3rdparty/windows
.TDatums
used for include/openpose/wrapper/wrapper.hpp
has changed from std::vector<Datum>
to std::vector<std::shared_ptr<Datum>>
, including the respective changes in all the worker classes. In addition, some template classes have been simplified to only take 1 template parameter for user simplicity.--write_coco_foot_json
by --write_coco_json_variants
in order to generalize to any COCO JSON format (i.e., hand, face, etc).scaleKeypoints2d
was not working if any of the scales was 1 (e.g., fail if scaleX = 1 but scaleY != 1, or if any offset was not 0).KeepTopNPeople
that could provoke segmentation fault for number_people_max
> 1./
(e.g., ~/Desktop/
worked but ~/Desktop
did not).--hand_alpha_pose
.process()
in bodyPartConnectorBase.cu
to defaultNmsThreshold
. This also removes any posibility of future bugs in that function for using a default NMS threshold higher than 0.15 (which was the hard-coded value used previously).removePeopleBelowThresholdsAndFillFaces
is re-run with maximizePositives = true
.--3d_min_views
default value (-1) no longer means that all camera views are required. Instead, it will be equal to max(2, min(4, #cameras-1)). This should provide a good trade-off between recall and precission.06_face_from_image.cpp
, 07_hand_from_image.cpp
, and 09_keypoints_from_heatmaps
working again, they stopped working in version 1.5.0 with the GPU image resize for the GUI.WrapperT::waitAndEmplace()
and WrapperT::tryEmplace()
improved, allowing multi-camera/3-D (TDatums
of size > 1).createMultiviewTDatum()
to auto-generate a TDatums
for multi-camera/3-D from a single cv::Mat (that is splitted) and the desired camera parameter matrices.examples/tutorial_api_cpp/11_asynchronous_custom_input_multi_camera.cpp
for a test example.dll
files in Windows, so
in Ubuntu) are still required.openpose_private
directory with some internal headers that, if exported with OpenPose, would require including 3rd-party headers and static library files.Releases
, subsection OpenCV - 4.2.0
, Windows
version.*.cpp
files, their include of their analog *.hpp
file has been moved to the first line of those *.cpp
files to slightly speed up compiling time.include/openpose/wrapper/
to avoid std::string to cause errors for using diferent std DLLs.ScaleMode::ZeroToOneFixedAspect
and ScaleMode::PlusMinusOneFixedAspect
. Compared to ZeroToOne
and PlusMinusOne
, the new ones also preserve the aspect ratio of each axis.openpose_private
, all 3rd-party library calls in headers, and std::string calls in include/openpose/wrapper/
.dLog()
as opLogIfDebug()
, log()
as opLog()
, check()
as checkBool()
, and also renamed all the checkX()
functions in include/openpose/utilities/check.hpp
. This avoids compiling crashes when exporting OpenPose to other projects which contain other 3rd-party libraries that define functions with the same popular names with #define
.output_resolution
was not working with GPU resize, redirected to CPU in those cases.Download and/or check any OpenPose version from https://github.com/CMU-Perceptual-Computing-Lab/openpose/releases.
此处可能存在不合适展示的内容,页面不予展示。您可通过相关编辑功能自查并修改。
如您确认内容无涉及 不当用语 / 纯广告导流 / 暴力 / 低俗色情 / 侵权 / 盗版 / 虚假 / 无价值内容或违法国家有关法律法规的内容,可点击提交进行申诉,我们将尽快为您处理。