Everything works properly

Pose2Sim/Demo/User/Config.toml

@@ -5,7 +5,7 @@
 # Configure your project parameters here
 
 [project]
-project_dir = 'D:\softs\github_david\pose2sim\Pose2Sim\Demo' # BETWEEN SINGLE QUOTES! # If empty, project dir is current dir
+project_dir = '' # BETWEEN SINGLE QUOTES! # If empty, project dir is current dir
 frame_range = [] #For example [10,300], or [] for all frames
 frame_rate = 60 #Hz
 
@@ -26,7 +26,7 @@ pose_model = 'BODY_25B' #CUSTOM, BODY_25B, BODY_25, BODY_135, BLAZEPOSE, HALPE_2
 
 
 [calibration]
-calibration_type = 'calculate' # 'convert' or 'calculate'
+calibration_type = 'convert' # 'convert' or 'calculate'
 
    [calibration.convert]
    convert_from = 'qualisys' # 'qualisys', 'optitrack', or 'vicon'
@@ -41,10 +41,9 @@ calibration_type = 'calculate' # 'convert' or 'calculate'
       intrinsics_board_type = 'checkerboard' # 'checkerboard' ('charucoboard' not supported yet)
       overwrite_intrinsics = false # overwrite (or not) if they have already been calculated?
       show_detection_intrinsics = true # true or false (lowercase)
-
       intrinsics_extension = 'jpg' # any video or image extension
-      extract_every_N_sec = 1 # if video, extract frames every N seconds (can be <1 )
 
+      extract_every_N_sec = 1 # if video, extract frames every N seconds (can be <1 )
       intrinsics_corners_nb = [4,7] 
       intrinsics_square_size = 60 # mm
       intrinsics_marker_size = 40 # mm # only checked if charucoboard
@@ -56,7 +55,6 @@ calibration_type = 'calculate' # 'convert' or 'calculate'
       # 'scene' involves manually clicking any point of know coordinates on scene. Usually more accurate if points are spread out
       calculate_extrinsics = true # true or false (lowercase)
       show_reprojection_error = true # true or false (lowercase)
-      
       extrinsics_extension = 'png' # any video or image extension
       
       # if extrinsics_board_type = 'checkerboard' or 'charucoboard'
@@ -87,7 +85,8 @@ likelihood_threshold = 0.3
 min_cameras_for_triangulation = 2
 interpolation = 'cubic' #linear, slinear, quadratic, cubic, or none
 # 'none' if you don't want to interpolate missing points
-show_interp_indices = false # true or false (lowercase). For each keypoint, return the frames that need to be interpolated
+interp_if_gap_smaller_than = 10 # do not interpolate bigger gaps
+show_interp_indices = true # true or false (lowercase). For each keypoint, return the frames that need to be interpolated
 
 
 [filtering]

Pose2Sim/Demo/User/test.toml

@@ -1,5 +1,5 @@
 ###############################################################################
-## POSE2SIM PROJECT PARAMETERS                                               ##
+## PROJECT PARAMETERS                                                        ##
 ###############################################################################
 
 # Configure your project parameters here
@@ -9,45 +9,78 @@ project_dir = '' # BETWEEN SINGLE QUOTES! # If empty, project dir is current dir
 frame_range = [0,10] #For example [10,300], or [] for all frames
 frame_rate = 60 #Hz
 
-rawImg_folder_name = 'raw-2d'
+rawImg_folder_name = 'raw'
 calib_folder_name = 'calibration'
 pose_folder_name = 'pose'
 pose_json_folder_extension = 'json'
 pose_img_folder_extension = 'img'
-poseTracked_folder_name = 'pose-2d-tracked'
+poseAssociated_folder_name = 'pose-associated'
-pose3d_folder_name = 'triangulation'
+pose3d_folder_name = 'pose-3d'
 opensim_folder_name = 'opensim'
 
+
 [pose]
-pose_model = 'BODY_25B' #CUSTOM, BODY_25B, BODY_25, BODY_135, BLAZEPOSE, 
+pose_model = 'BODY_25B' #CUSTOM, BODY_25B, BODY_25, BODY_135, BLAZEPOSE, HALPE_26, HALPE_68, HALPE_136, COCO_133, COCO, MPII are available,
-# HALPE_26, HALPE_68, HALPE_136, COCO_133, COCO, MPII are available,
 # from DeepLabCut, OpenPose, MediaPipe BlazePose, and AlphaPose
 # See Pose2Sim\skeleton.py for their skeleton hierarchy
 
+
 [calibration]
-type = 'qca' # 'qca', 'checkerboard', 'arucoboard', or 'charucoboard'
+calibration_type = 'convert' # 'convert' or 'calculate'
-   [calibration.qca]
-   binning_factor = 1 # Usually 1, except when filming in 540p where it usually is 2
 
-   [calibration.checkerboard]
+   [calibration.convert]
-   corners_nb = [7,12] # [H,W] rather than [w,h]
+   convert_from = 'qualisys' # 'qualisys', 'optitrack', or 'vicon'
-   square_size = 80 # mm # [h,w] if square is actually a rectangle
+      [calibration.convert.qualisys]
-   frame_for_origin = -1 # starting from zero. -1 if board is at origin on last frame
+      binning_factor = 1 # Usually 1, except when filming in 540p where it usually is 2
-   # /!\ Beware that corners must be detected on all view at frame_for_origin, or else 
-   # extrinsic parameters will be wrong. Set show_corner_detection to true to check it.
-   show_corner_detection = false # true or false (lowercase)
-   from_vid_or_img = 'img' # 'vid' or 'img'
-   vid_snapshot_every_N_frames = 20
-   vid_extension = 'mp4'
-   img_extension = 'jpg' # 'png', 'jpg', etc
   
-[2d-tracking]
+
+   [calibration.calculate] 
+   calculate_method = 'board' # 'board' or 'points'
+      
+      [calibration.calculate.board.intrinsics] # camera properties, only needs to be done once
+      intrinsics_board_type = 'checkerboard' # 'checkerboard' ('charucoboard' not supported yet)
+      overwrite_intrinsics = false # overwrite (or not) if they have already been calculated?
+      show_detection_intrinsics = true # true or false (lowercase)
+      intrinsics_extension = 'jpg' # any video or image extension
+
+      extract_every_N_sec = 1 # if video, extract frames every N seconds (can be <1 )
+      intrinsics_corners_nb = [4,7] 
+      intrinsics_square_size = 60 # mm
+      intrinsics_marker_size = 40 # mm # only checked if charucoboard
+      intrinsics_aruco_dict = 'DICT_6X6_250' # only checked if charucoboard # see https://docs.opencv.org/3.4/dc/df7/dictionary_8hpp.html
+   
+      [calibration.calculate.board.extrinsics] # camera placement, needs to be done every time
+      extrinsics_board_type = 'scene' # 'checkerboard', 'scene'  ('charucoboard' not supported yet)
+      # 'board' should be large enough to be detected when laid on the floor. 
+      # 'scene' involves manually clicking any point of know coordinates on scene. Usually more accurate if points are spread out
+      calculate_extrinsics = true # true or false (lowercase)
+      show_reprojection_error = true # true or false (lowercase)
+      extrinsics_extension = 'png' # any video or image extension
+      
+      # if extrinsics_board_type = 'checkerboard' or 'charucoboard'
+      extrinsics_corners_nb = [4,7] # [H,W] rather than [w,h]
+      extrinsics_square_size = 60 # mm # [h,w] if square is actually a rectangle
+      extrinsics_marker_size = 40 # mm # only checked if 'charucoboard' (not supported yet)
+      extrinsics_aruco_dict = 'DICT_6X6_250' # only checked if 'charucoboard' # see https://docs.opencv.org/3.4/dc/df7/dictionary_8hpp.html
+      
+      # if extrinsics_board_type = 'scene'
+      # list of 3D coordinates to be manually labelled on images. Can also be a 2 dimensional plane. # in m
+      object_coords_3d = [[-2.0,  0.3,  0.0], [-2.0 , 0.0,  0.0], [-2.0, 0.0,  0.05], [-2.0, -0.3 ,  0.0], [0.0,  0.3,  0.0], [0.0, 0.0,  0.0], [0.0, 0.0,  0.05], [0.0, -0.3,  0.0]] # in meters -> Not in mm! <-
+      
+  
+      [calibration.calculate.points]
+      calibration_points = 'wand' # 'wand' or 'keypoints'
+      # Not supported yet.
+
+	  
+[personAssociation]
 tracked_keypoint = 'Neck' # If the neck is not detected by the pose_model, check skeleton.py 
 # and choose a stable point for tracking the person of interest (e.g., 'right_shoulder' with BLAZEPOSE)
-error_threshold_tracking = 20 # px
+reproj_error_threshold_association = 20 # px
 
-[3d-triangulation]
+
-error_threshold_triangulation = 15 # px
+[triangulation]
+reproj_error_threshold_triangulation = 15 # px
 likelihood_threshold = 0.3
 min_cameras_for_triangulation = 2
 interpolation = 'cubic' #linear, slinear, quadratic, cubic, or none
@@ -55,23 +88,26 @@ interpolation = 'cubic' #linear, slinear, quadratic, cubic, or none
 interp_if_gap_smaller_than = 10 # do not interpolate bigger gaps
 show_interp_indices = true # true or false (lowercase). For each keypoint, return the frames that need to be interpolated
 
-[3d-filtering]
-type = 'butterworth' # butterworth, butterworth_on_speed, gaussian, LOESS, median
-display_figures = false # true or false (lowercase)
 
-   [3d-filtering.butterworth]
+[filtering]
-   type = 'low'
+type = 'butterworth' # butterworth, kalman, gaussian, LOESS, median, butterworth_on_speed
+display_figures = true # true or false (lowercase)
+
+   [filtering.butterworth]
    order = 4 
    cut_off_frequency = 6 # Hz
-   [3d-filtering.butterworth_on_speed]
+   [filtering.kalman]
-   type = 'low'
+   # How much more do you trust triangulation results (measurements), than previous data (process assuming constant acceleration)?
+   trust_ratio = 100 # = measurement_trust/process_trust ~= process_noise/measurement_noise
+   smooth = true # should be true, unless you need real-time filtering
+   [filtering.butterworth_on_speed]
    order = 4 
    cut_off_frequency = 10 # Hz
-   [3d-filtering.gaussian]
+   [filtering.gaussian]
    sigma_kernel = 2 #px
-   [3d-filtering.LOESS]
+   [filtering.LOESS]
-   nb_values_used = 5 # = fraction of data used * nb frames
+   nb_values_used = 30 # = fraction of data used * nb frames
-   [3d-filtering.median]
+   [filtering.median]
    kernel_size = 9
 
 [opensim]

Pose2Sim/Demo/calibration/Calib_intcheckerboard_extscene.toml

@@ -1,39 +0,0 @@
-[cam_01]
-name = "cam_01"
-size = [ 1088.0, 1920.0]
-matrix = [ [ 1681.244873046875, 0.0, 532.97369384375], [ 0.0, 1681.075439453125, 948.137390140625], [ 0.0, 0.0, 1.0]]
-distortions = [ -0.000721609375, 0.002187234375, 9.5e-06, 1.078125e-05]
-rotation = [ [1.70179346], [1.04839803], [-0.40602443]]
-translation = [ [0.31962238], [0.93995713], [2.97508412]]
-fisheye = false
-
-[cam_02]
-name = "cam_02"
-size = [ 1088.0, 1920.0]
-matrix = [ [ 1673.729614265625, 0.0, 534.494567875], [ 0.0, 1673.79724121875, 963.225891109375], [ 0.0, 0.0, 1.0]]
-distortions = [ -0.000747609375, 0.00213728125, 1.51875e-05, 4.546875e-06]
-rotation = [ [1.35902244], [1.60141928], [-1.18177902]]
-translation = [ [-0.12599186], [0.74502948], [3.17221429]]
-fisheye = false
-
-[cam_03]
-name = "cam_03"
-size = [ 1088.0, 1920.0]
-matrix = [ [ 1681.598388671875, 0.0, 513.20837403125], [ 0.0, 1681.509887703125, 955.005126953125], [ 0.0, 0.0, 1.0]]
-distortions = [ -0.000729765625, 0.00215034375, -8.46875e-06, -8.078125e-06]
-rotation = [ [0.81816165], [-2.22184308], [1.4608978]]
-translation = [ [-0.76301979], [0.34503189], [4.43845847]]
-fisheye = false
-
-[cam_04]
-name = "cam_04"
-size = [ 1088.0, 1920.0]
-matrix = [ [ 1675.234985359375, 0.0, 540.106201171875], [ 0.0, 1675.204223640625, 964.0302734375], [ 0.0, 0.0, 1.0]]
-distortions = [ -0.000744265625, 0.002104171875, 4.328125e-06, 3.109375e-06]
-rotation = [ [1.4245729], [-1.39512301], [0.45721172]]
-translation = [ [0.48659282], [0.02675454], [4.33496911]]
-fisheye = false
-
-[metadata]
-adjusted = false
-error = 0.0

Pose2Sim/Demo/calibration/Calib_intcheckerboard_extzeros.toml.old

@@ -1,39 +0,0 @@
-[cam_1]
-name = "cam_01"
-size = [ 1088.0, 1920.0]
-matrix = [ [ 1687.6396622414502, 0.0, 543.5], [ 0.0, 1686.6625452424132, 959.5], [ 0.0, 0.0, 1.0]]
-distortions = [ -0.05054882348909584, 0.1576374662929963, 0.002210439026715273, -0.002164250840100675]
-rotation = [ 0, 0, 0]
-translation = [ 0, 0, 0]
-fisheye = false
-
-[cam_2]
-name = "cam_02"
-size = [ 1088.0, 1920.0]
-matrix = [ [ 1687.859605173682, 0.0, 543.5], [ 0.0, 1685.4371774565182, 959.5], [ 0.0, 0.0, 1.0]]
-distortions = [ -0.03518095956993663, 0.037170387967130306, 0.0017983248726869015, -0.0018747264134721386]
-rotation = [ 0, 0, 0]
-translation = [ 0, 0, 0]
-fisheye = false
-
-[cam_3]
-name = "cam_03"
-size = [ 1088.0, 1920.0]
-matrix = [ [ 1680.286039024573, 0.0, 543.5], [ 0.0, 1681.1761700950424, 959.5], [ 0.0, 0.0, 1.0]]
-distortions = [ -0.030304789365317657, 0.04613237843981006, 0.004618104732402956, -0.001684974310757026]
-rotation = [ 0, 0, 0]
-translation = [ 0, 0, 0]
-fisheye = false
-
-[cam_4]
-name = "cam_04"
-size = [ 1088.0, 1920.0]
-matrix = [ [ 1689.483288223434, 0.0, 543.5], [ 0.0, 1688.5630322142706, 959.5], [ 0.0, 0.0, 1.0]]
-distortions = [ -0.04369424846208241, 0.12212413289800295, 0.0017367557051414114, -0.0018985222378279064]
-rotation = [ 0, 0, 0]
-translation = [ 0, 0, 0]
-fisheye = false
-
-[metadata]
-adjusted = false
-error = 0.0

Pose2Sim/Demo/calibration/Calib_qualisys.toml

@@ -1,39 +0,0 @@
-[cam_01]
-name = "cam_01"
-size = [ 1088.0, 1920.0]
-matrix = [ [ 1681.244873046875, 0.0, 532.97369384375], [ 0.0, 1681.075439453125, 948.137390140625], [ 0.0, 0.0, 1.0]]
-distortions = [ -0.000721609375, 0.002187234375, 9.5e-06, 1.078125e-05]
-rotation = [ 1.6882754799999993, 1.0483220499999997, -0.41955852000000016]
-translation = [ 0.3211048899999996, 0.9563320600000009, 2.8907130499999996]
-fisheye = false
-
-[cam_02]
-name = "cam_02"
-size = [ 1088.0, 1920.0]
-matrix = [ [ 1673.729614265625, 0.0, 534.494567875], [ 0.0, 1673.79724121875, 963.225891109375], [ 0.0, 0.0, 1.0]]
-distortions = [ -0.000747609375, 0.00213728125, 1.51875e-05, 4.546875e-06]
-rotation = [ 1.34975875, 1.5963809099999993, -1.1983285799999999]
-translation = [ -0.11152829000000017, 0.7766184800000001, 3.0675519599999994]
-fisheye = false
-
-[cam_03]
-name = "cam_03"
-size = [ 1088.0, 1920.0]
-matrix = [ [ 1681.598388671875, 0.0, 513.20837403125], [ 0.0, 1681.509887703125, 955.005126953125], [ 0.0, 0.0, 1.0]]
-distortions = [ -0.000729765625, 0.00215034375, -8.46875e-06, -8.078125e-06]
-rotation = [ 0.8109654899999995, -2.1972129299999996, 1.3760277799999996]
-translation = [ -0.7934803899999996, 0.32283594000000126, 4.353514870000001]
-fisheye = false
-
-[cam_04]
-name = "cam_04"
-size = [ 1088.0, 1920.0]
-matrix = [ [ 1675.234985359375, 0.0, 540.106201171875], [ 0.0, 1675.204223640625, 964.0302734375], [ 0.0, 0.0, 1.0]]
-distortions = [ -0.000744265625, 0.002104171875, 4.328125e-06, 3.109375e-06]
-rotation = [ 1.4045571699999995, -1.3887412699999993, 0.42535743000000026]
-translation = [ 0.5030217200000007, 0.04894934000000083, 4.406564460000002]
-fisheye = false
-
-[metadata]
-adjusted = false
-error = 0.0

Pose2Sim/Demo/calibration/Calib_qualisys.toml.old

@@ -1,39 +0,0 @@
-[cam_1]
-name = "cam01"
-size = [ 1088.0, 1920.0]
-matrix = [ [ 1681.244873046875, 0.0, 532.97369384375], [ 0.0, 1681.075439453125, 948.137390140625], [ 0.0, 0.0, 1.0]]
-distortions = [ -0.000721609375, 0.002187234375, 9.5e-06, 1.078125e-05]
-rotation = [ 1.6882754799999993, 1.0483220499999997, -0.41955852000000016]
-translation = [ 0.3211048899999996, 0.9563320600000009, 2.8907130499999996]
-fisheye = false
-
-[cam_2]
-name = "cam02"
-size = [ 1088.0, 1920.0]
-matrix = [ [ 1673.729614265625, 0.0, 534.494567875], [ 0.0, 1673.79724121875, 963.225891109375], [ 0.0, 0.0, 1.0]]
-distortions = [ -0.000747609375, 0.00213728125, 1.51875e-05, 4.546875e-06]
-rotation = [ 1.34975875, 1.5963809099999993, -1.1983285799999999]
-translation = [ -0.11152829000000017, 0.7766184800000001, 3.0675519599999994]
-fisheye = false
-
-[cam_3]
-name = "cam03"
-size = [ 1088.0, 1920.0]
-matrix = [ [ 1681.598388671875, 0.0, 513.20837403125], [ 0.0, 1681.509887703125, 955.005126953125], [ 0.0, 0.0, 1.0]]
-distortions = [ -0.000729765625, 0.00215034375, -8.46875e-06, -8.078125e-06]
-rotation = [ 0.8109654899999995, -2.1972129299999996, 1.3760277799999996]
-translation = [ -0.7934803899999996, 0.32283594000000126, 4.353514870000001]
-fisheye = false
-
-[cam_4]
-name = "cam04"
-size = [ 1088.0, 1920.0]
-matrix = [ [ 1675.234985359375, 0.0, 540.106201171875], [ 0.0, 1675.204223640625, 964.0302734375], [ 0.0, 0.0, 1.0]]
-distortions = [ -0.000744265625, 0.002104171875, 4.328125e-06, 3.109375e-06]
-rotation = [ 1.4045571699999995, -1.3887412699999993, 0.42535743000000026]
-translation = [ 0.5030217200000007, 0.04894934000000083, 4.406564460000002]
-fisheye = false
-
-[metadata]
-adjusted = false
-error = 0.0

Pose2Sim/Empty_project/User/Config.toml

@@ -1,5 +1,5 @@
 ###############################################################################
-## POSE2SIM PROJECT PARAMETERS                                               ##
+## PROJECT PARAMETERS                                                        ##
 ###############################################################################
 
 # Configure your project parameters here
@@ -9,45 +9,78 @@ project_dir = '' # BETWEEN SINGLE QUOTES! # If empty, project dir is current dir
 frame_range = [] #For example [10,300], or [] for all frames
 frame_rate = 60 #Hz
 
-rawImg_folder_name = 'raw-2d'
+rawImg_folder_name = 'raw'
 calib_folder_name = 'calibration'
 pose_folder_name = 'pose'
 pose_json_folder_extension = 'json'
 pose_img_folder_extension = 'img'
-poseTracked_folder_name = 'pose-2d-tracked'
+poseAssociated_folder_name = 'pose-associated'
-pose3d_folder_name = 'triangulation'
+pose3d_folder_name = 'pose-3d'
 opensim_folder_name = 'opensim'
 
+
 [pose]
-pose_model = 'BODY_25B' #CUSTOM, BODY_25B, BODY_25, BODY_135, BLAZEPOSE, 
+pose_model = 'BODY_25B' #CUSTOM, BODY_25B, BODY_25, BODY_135, BLAZEPOSE, HALPE_26, HALPE_68, HALPE_136, COCO_133, COCO, MPII are available,
-# HALPE_26, HALPE_68, HALPE_136, COCO_133, COCO, MPII are available,
 # from DeepLabCut, OpenPose, MediaPipe BlazePose, and AlphaPose
 # See Pose2Sim\skeleton.py for their skeleton hierarchy
 
+
 [calibration]
-type = 'qca' # 'qca', 'checkerboard', 'arucoboard', or 'charucoboard'
+calibration_type = 'convert' # 'convert' or 'calculate'
-   [calibration.qca]
-   binning_factor = 1 # Usually 1, except when filming in 540p where it usually is 2
 
-   [calibration.checkerboard]
+   [calibration.convert]
-   corners_nb = [7,12] # [H,W] rather than [w,h]
+   convert_from = 'qualisys' # 'qualisys', 'optitrack', or 'vicon'
-   square_size = 80 # mm # [h,w] if square is actually a rectangle
+      [calibration.convert.qualisys]
-   frame_for_origin = -1 # starting from zero. -1 if board is at origin on last frame
+      binning_factor = 1 # Usually 1, except when filming in 540p where it usually is 2
-   # /!\ Beware that corners must be detected on all view at frame_for_origin, or else 
-   # extrinsic parameters will be wrong. Set show_corner_detection to true to check it.
-   show_corner_detection = false # true or false (lowercase)
-   from_vid_or_img = 'img' # 'vid' or 'img'
-   vid_snapshot_every_N_frames = 20
-   vid_extension = 'mp4'
-   img_extension = 'jpg' # 'png', 'jpg', etc
   
-[2d-tracking]
+
+   [calibration.calculate] 
+   calculate_method = 'board' # 'board' or 'points'
+      
+      [calibration.calculate.board.intrinsics] # camera properties, only needs to be done once
+      intrinsics_board_type = 'checkerboard' # 'checkerboard' ('charucoboard' not supported yet)
+      overwrite_intrinsics = false # overwrite (or not) if they have already been calculated?
+      show_detection_intrinsics = true # true or false (lowercase)
+      intrinsics_extension = 'jpg' # any video or image extension
+
+      extract_every_N_sec = 1 # if video, extract frames every N seconds (can be <1 )
+      intrinsics_corners_nb = [4,7] 
+      intrinsics_square_size = 60 # mm
+      intrinsics_marker_size = 40 # mm # only checked if charucoboard
+      intrinsics_aruco_dict = 'DICT_6X6_250' # only checked if charucoboard # see https://docs.opencv.org/3.4/dc/df7/dictionary_8hpp.html
+   
+      [calibration.calculate.board.extrinsics] # camera placement, needs to be done every time
+      extrinsics_board_type = 'scene' # 'checkerboard', 'scene'  ('charucoboard' not supported yet)
+      # 'board' should be large enough to be detected when laid on the floor. 
+      # 'scene' involves manually clicking any point of know coordinates on scene. Usually more accurate if points are spread out
+      calculate_extrinsics = true # true or false (lowercase)
+      show_reprojection_error = true # true or false (lowercase)
+      extrinsics_extension = 'png' # any video or image extension
+      
+      # if extrinsics_board_type = 'checkerboard' or 'charucoboard'
+      extrinsics_corners_nb = [4,7] # [H,W] rather than [w,h]
+      extrinsics_square_size = 60 # mm # [h,w] if square is actually a rectangle
+      extrinsics_marker_size = 40 # mm # only checked if 'charucoboard' (not supported yet)
+      extrinsics_aruco_dict = 'DICT_6X6_250' # only checked if 'charucoboard' # see https://docs.opencv.org/3.4/dc/df7/dictionary_8hpp.html
+      
+      # if extrinsics_board_type = 'scene'
+      # list of 3D coordinates to be manually labelled on images. Can also be a 2 dimensional plane. # in m
+      object_coords_3d = [[-2.0,  0.3,  0.0], [-2.0 , 0.0,  0.0], [-2.0, 0.0,  0.05], [-2.0, -0.3 ,  0.0], [0.0,  0.3,  0.0], [0.0, 0.0,  0.0], [0.0, 0.0,  0.05], [0.0, -0.3,  0.0]] # in meters -> Not in mm! <-
+      
+  
+      [calibration.calculate.points]
+      calibration_points = 'wand' # 'wand' or 'keypoints'
+      # Not supported yet.
+
+	  
+[personAssociation]
 tracked_keypoint = 'Neck' # If the neck is not detected by the pose_model, check skeleton.py 
 # and choose a stable point for tracking the person of interest (e.g., 'right_shoulder' with BLAZEPOSE)
-error_threshold_tracking = 20 # px
+reproj_error_threshold_association = 20 # px
 
-[3d-triangulation]
+
-error_threshold_triangulation = 15 # px
+[triangulation]
+reproj_error_threshold_triangulation = 15 # px
 likelihood_threshold = 0.3
 min_cameras_for_triangulation = 2
 interpolation = 'cubic' #linear, slinear, quadratic, cubic, or none
@@ -55,23 +88,26 @@ interpolation = 'cubic' #linear, slinear, quadratic, cubic, or none
 interp_if_gap_smaller_than = 10 # do not interpolate bigger gaps
 show_interp_indices = true # true or false (lowercase). For each keypoint, return the frames that need to be interpolated
 
-[3d-filtering]
+
-type = 'butterworth' # butterworth, butterworth_on_speed, gaussian, LOESS, median
+[filtering]
+type = 'butterworth' # butterworth, kalman, gaussian, LOESS, median, butterworth_on_speed
 display_figures = true # true or false (lowercase)
 
-   [3d-filtering.butterworth]
+   [filtering.butterworth]
-   type = 'low'
    order = 4 
    cut_off_frequency = 6 # Hz
-   [3d-filtering.butterworth_on_speed]
+   [filtering.kalman]
-   type = 'low'
+   # How much more do you trust triangulation results (measurements), than previous data (process assuming constant acceleration)?
+   trust_ratio = 100 # = measurement_trust/process_trust ~= process_noise/measurement_noise
+   smooth = true # should be true, unless you need real-time filtering
+   [filtering.butterworth_on_speed]
    order = 4 
    cut_off_frequency = 10 # Hz
-   [3d-filtering.gaussian]
+   [filtering.gaussian]
    sigma_kernel = 2 #px
-   [3d-filtering.LOESS]
+   [filtering.LOESS]
-   nb_values_used = 5 # = fraction of data used * nb frames
+   nb_values_used = 30 # = fraction of data used * nb frames
-   [3d-filtering.median]
+   [filtering.median]
    kernel_size = 9
 
 [opensim]

Pose2Sim/calibration.py

@@ -136,7 +136,7 @@ def read_qca(qca_path, binning_factor):
     
     # Camera name
     for i, tag in enumerate(root.findall('cameras/camera')):
-        ret += [float(tag.attrib.get('avg-residual'))/1000]
+        ret += [float(tag.attrib.get('avg-residual'))]
         C += [tag.attrib.get('serial')]
         if tag.attrib.get('model') in ('Miqus Video', 'Miqus Video UnderWater', 'none'):
             vid_id += [i]
@@ -423,15 +423,13 @@ def calibrate_intrinsics(calib_dir, intrinsics_config_dict):
             logging.exception(f'The folder {os.path.join(calib_dir, "intrinsics", cam)} does not exist or does not contain any files with extension .{intrinsics_extension}.')
             raise ValueError(f'The folder {os.path.join(calib_dir, "intrinsics", cam)} does not exist or does not contain any files with extension .{intrinsics_extension}.')
         img_vid_files = sorted(img_vid_files, key=lambda c: [int(n) for n in re.findall(r'\d+', c)]) #sorting paths with numbers
+        
         # extract frames from video if video
         try:
-            # check if file is a video rather than an image
             cap = cv2.VideoCapture(img_vid_files[0])
             cap.read()
             if cap.read()[0] == False:
-                logging.exception('No video in the folder or wrong extension.')
+                raise
-                raise ValueError('No video in the folder or wrong extension.')
-            # extract frames from video
             extract_frames(img_vid_files[0], extract_every_N_sec, overwrite_extraction)
             img_vid_files = glob.glob(os.path.join(calib_dir, 'intrinsics', cam, f'*.png'))
             img_vid_files = sorted(img_vid_files, key=lambda c: [int(n) for n in re.findall(r'\d+', c)])
@@ -442,16 +440,18 @@ def calibrate_intrinsics(calib_dir, intrinsics_config_dict):
         for img_path in img_vid_files:
             if show_detection_intrinsics == True:
                 imgp_confirmed, objp_confirmed = findCorners(img_path, intrinsics_corners_nb, objp=objp, show=show_detection_intrinsics)
+                print(len(imgp_confirmed), len(objp_confirmed))
                 if isinstance(imgp_confirmed, np.ndarray):
                     imgpoints.append(imgp_confirmed)
                     objpoints.append(objp_confirmed)
+
             else:
                 imgp_confirmed = findCorners(img_path, intrinsics_corners_nb, objp=objp, show=show_detection_intrinsics)
                 if isinstance(imgp_confirmed, np.ndarray):
                     imgpoints.append(imgp_confirmed)
                     objpoints.append(objp)
         if len(imgpoints) < 10:
-            logging.info(f'Corners were detected only on {len(imgpoints)} images for camera {cam}. Calibration of intrinsic parameters may not be accurate with less than 10 good images of the board.')
+            logging.info(f'Corners were detected only on {len(imgpoints)} images for camera {cam}. Calibration of intrinsic parameters may not be accurate with fewer than 10 good images of the board.')
 
         # calculate intrinsics
         img = cv2.imread(str(img_path))
@@ -524,11 +524,12 @@ def calibrate_extrinsics(calib_dir, extrinsics_config_dict, C, S, K, D):
                 logging.exception('No points clicked (or fewer than 6). Press \'C\' when the image is displayed, and then click on the image points corresponding to the \'object_coords_3d\' you measured and wrote down in the Config.toml file.')
                 raise ValueError('No points clicked (or fewer than 6). Press \'C\' when the image is displayed, and then click on the image points corresponding to the \'object_coords_3d\' you measured and wrote down in the Config.toml file.')
             if len(objp) < 10:
-                logging.info(f'Only {len(objp)} reference points for camera {cam}. Calibration of extrinsic parameters may not be accurate with less than 10 reference points, as spread out as possible.')
+                logging.info(f'Only {len(objp)} reference points for camera {cam}. Calibration of extrinsic parameters may not be accurate with fewer than 10 reference points, as spread out as possible.')
 
         # Calculate extrinsics
         mtx, dist = np.array(K[i]), np.array(D[i])
         _, r, t = cv2.solvePnP(objp, imgp, mtx, dist)
+        r, t = r.flatten(), t.flatten()
 
         # Projection of object points to image plane
         Kh_cam = np.block([mtx, np.zeros(3).reshape(3,1)])
@@ -545,9 +546,16 @@ def calibrate_extrinsics(calib_dir, extrinsics_config_dict, C, S, K, D):
                 cv2.circle(img, (int(o[0]), int(o[1])), 8, (0,0,255), -1) 
             for i in imgp:
                 cv2.drawMarker(img, (int(i[0][0]), int(i[0][1])), (0,255,0), cv2.MARKER_CROSS, 15, 2)
+            cv2.putText(img, 'Verify calibration results, then close window.', (20, 20), cv2.FONT_HERSHEY_SIMPLEX, .5, (255,255,255), 2, lineType = cv2.LINE_AA)
+            cv2.putText(img, 'Verify calibration results, then close window.', (20, 20), cv2.FONT_HERSHEY_SIMPLEX, .5, (0,0,0), 1, lineType = cv2.LINE_AA) 
+            cv2.drawMarker(img, (20,40), (0,255,0), cv2.MARKER_CROSS, 15, 2)
+            cv2.putText(img, '    Clicked points', (20, 40), cv2.FONT_HERSHEY_SIMPLEX, .5, (255,255,255), 2, lineType = cv2.LINE_AA)
+            cv2.putText(img, '    Clicked points', (20, 40), cv2.FONT_HERSHEY_SIMPLEX, .5, (0,0,0), 1, lineType = cv2.LINE_AA)    
+            cv2.circle(img, (20,60), 8, (0,0,255), -1)    
+            cv2.putText(img, '    Reprojected object points', (20, 60), cv2.FONT_HERSHEY_SIMPLEX, .5, (255,255,255), 2, lineType = cv2.LINE_AA)
+            cv2.putText(img, '    Reprojected object points', (20, 60), cv2.FONT_HERSHEY_SIMPLEX, .5, (0,0,0), 1, lineType = cv2.LINE_AA)    
             im_pil = Image.fromarray(img)
-            print(img_vid_files[0])
+            im_pil.show(title = os.path.basename(img_vid_files[0]))
-            im_pil.show(title = img_vid_files[0])
 
         # Calculate reprojection error
         imgp_to_objreproj_dist = [euclidean_distance(proj_obj[n], imgp[n]) for n in range(len(proj_obj))]
@@ -574,7 +582,7 @@ def findCorners(img_path, corner_nb, objp=[], show=True):
     
     INPUTS:
     - img_path: path to image (or video)
-    - corner_nb: [H, W] internal corners in checkerboard: list of two integers [9,6]
+    - corner_nb: [H, W] internal corners in checkerboard: list of two integers [4,7]
     - optionnal: show: choose whether to show corner detections
     - optionnal: objp: array [3d corner coordinates]
 
@@ -678,7 +686,7 @@ def imgp_objp_visualizer_clicker(img, imgp=[], objp=[], img_path=''):
                 objp_confirmed = objp
             else:
                 imgp_confirmed = np.array([imgp.astype('float32') for imgp in imgp_confirmed])
-                objp_confirmed = np.array(objp_confirmed)
+                objp_confirmed = objp_confirmed
             # OpenCV needs at leas 4 correspondance points to calibrate
             if len(imgp_confirmed) < 6:
                 objp_confirmed = []
@@ -696,6 +704,9 @@ def imgp_objp_visualizer_clicker(img, imgp=[], objp=[], img_path=''):
             objp_confirmed = []
 
         if event.key == 'c':
+            # TODO: RIGHT NOW, IF 'C' IS PRESSED ANOTHER TIME, OBJP_CONFIRMED AND IMGP_CONFIRMED ARE RESET TO []
+            if 'objp_confirmed' in globals():
+                del objp_confirmed
             # If 'c', allows retrieving imgp_confirmed by clicking them on the image
             scat = ax.scatter([],[],s=100,marker='+',color='g')
             plt.connect('button_press_event', on_click)
@@ -731,7 +742,7 @@ def imgp_objp_visualizer_clicker(img, imgp=[], objp=[], img_path=''):
                     fig_3d.canvas.draw()
                 elif count == len(objp)-1:
                     # if all objp have been clicked or indicated as not visible, close all
-                    imgp_confirmed = np.array(imgp)
+                    imgp_confirmed = np.array(imgp, np.float32)
                     plt.close('all')
                     for var_to_delete in ['events', 'count', 'scat', 'fig_3d', 'ax_3d', 'objp_confirmed_notok']:
                         if var_to_delete in globals():
@@ -780,8 +791,8 @@ def imgp_objp_visualizer_clicker(img, imgp=[], objp=[], img_path=''):
                     fig_3d.canvas.draw()
                 elif count == len(objp)-1:
                     # retrieve objp_confirmed
-                    objp_confirmed = np.array([[objp[count].tolist()] if 'objp_confirmed' not in globals() else objp_confirmed+[objp[count]]][0])
+                    objp_confirmed = np.array([[objp[count]] if 'objp_confirmed' not in globals() else objp_confirmed+[objp[count]]][0])
-                    imgp_confirmed = np.array(imgp_confirmed)
+                    imgp_confirmed = np.array(imgp_confirmed, np.float32)
                     # close all, delete all
                     plt.close('all')
                     for var_to_delete in ['events', 'count', 'scat', 'scat_3d', 'fig_3d', 'ax_3d', 'objp_confirmed_notok']:
@@ -792,6 +803,8 @@ def imgp_objp_visualizer_clicker(img, imgp=[], objp=[], img_path=''):
                     objp_confirmed = [[objp[count]] if 'objp_confirmed' not in globals() else objp_confirmed+[objp[count]]][0]
                     ax_3d.scatter(*objp[count], marker='o', color='g')
                     fig_3d.canvas.draw()
+                print(objp_confirmed)
+                
 
         # Right click: 
         # If last event was left click, remove last point and if objp given, from objp_confirmed
@@ -858,20 +871,20 @@ def imgp_objp_visualizer_clicker(img, imgp=[], objp=[], img_path=''):
         ax.set_zlim3d([z_middle - plot_radius, z_middle + plot_radius])
 
     # Write instructions
-    cv2.putText(img, 'Type "Y" to accept point detection.', (20, 20), cv2.FONT_HERSHEY_SIMPLEX, .5, (255,255,255), 2, lineType = cv2.LINE_AA)
+    cv2.putText(img, 'Type "Y" to accept point detection.', (20, 20), cv2.FONT_HERSHEY_SIMPLEX, .7, (255,255,255), 3, lineType = cv2.LINE_AA)
-    cv2.putText(img, 'Type "Y" to accept point detection.', (20, 20), cv2.FONT_HERSHEY_SIMPLEX, .5, (0,0,0), 1, lineType = cv2.LINE_AA)    
+    cv2.putText(img, 'Type "Y" to accept point detection.', (20, 20), cv2.FONT_HERSHEY_SIMPLEX, .7, (0,0,0), 2, lineType = cv2.LINE_AA)    
-    cv2.putText(img, 'If points are wrongfully (or not) detected:', (20, 40), cv2.FONT_HERSHEY_SIMPLEX, .5, (255,255,255), 2, lineType = cv2.LINE_AA)
+    cv2.putText(img, 'If points are wrongfully (or not) detected:', (20, 43), cv2.FONT_HERSHEY_SIMPLEX, .7, (255,255,255), 3, lineType = cv2.LINE_AA)
-    cv2.putText(img, 'If points are wrongfully (or not) detected:', (20, 40), cv2.FONT_HERSHEY_SIMPLEX, .5, (0,0,0), 1, lineType = cv2.LINE_AA)    
+    cv2.putText(img, 'If points are wrongfully (or not) detected:', (20, 43), cv2.FONT_HERSHEY_SIMPLEX, .7, (0,0,0), 2, lineType = cv2.LINE_AA)    
-    cv2.putText(img, '- type "N" to dismiss this image,', (20, 60), cv2.FONT_HERSHEY_SIMPLEX, .5, (255,255,255), 2, lineType = cv2.LINE_AA)
+    cv2.putText(img, '- type "N" to dismiss this image,', (20, 66), cv2.FONT_HERSHEY_SIMPLEX, .7, (255,255,255), 3, lineType = cv2.LINE_AA)
-    cv2.putText(img, '- type "N" to dismiss this image,', (20, 60), cv2.FONT_HERSHEY_SIMPLEX, .5, (0,0,0), 1, lineType = cv2.LINE_AA)    
+    cv2.putText(img, '- type "N" to dismiss this image,', (20, 66), cv2.FONT_HERSHEY_SIMPLEX, .7, (0,0,0), 2, lineType = cv2.LINE_AA)    
-    cv2.putText(img, '- type "C" to click points by hand (beware of their order).', (20, 80), cv2.FONT_HERSHEY_SIMPLEX, .5, (255,255,255), 2, lineType = cv2.LINE_AA)
+    cv2.putText(img, '- type "C" to click points by hand (beware of their order).', (20, 89), cv2.FONT_HERSHEY_SIMPLEX, .7, (255,255,255), 3, lineType = cv2.LINE_AA)
-    cv2.putText(img, '- type "C" to click points by hand (beware of their order).', (20, 80), cv2.FONT_HERSHEY_SIMPLEX, .5, (0,0,0), 1, lineType = cv2.LINE_AA)    
+    cv2.putText(img, '- type "C" to click points by hand (beware of their order).', (20, 89), cv2.FONT_HERSHEY_SIMPLEX, .7, (0,0,0), 2, lineType = cv2.LINE_AA)    
-    cv2.putText(img, '   left click to add a point, right click to remove it, "H" to indicate it is not visible. ', (20, 100), cv2.FONT_HERSHEY_SIMPLEX, .5, (255,255,255), 2, lineType = cv2.LINE_AA)
+    cv2.putText(img, '   left click to add a point, right click to remove it, "H" to indicate it is not visible. ', (20, 112), cv2.FONT_HERSHEY_SIMPLEX, .7, (255,255,255), 3, lineType = cv2.LINE_AA)
-    cv2.putText(img, '   left click to add a point, right click to remove it, "H" to indicate it is not visible. ', (20, 100), cv2.FONT_HERSHEY_SIMPLEX, .5, (0,0,0), 1, lineType = cv2.LINE_AA)    
+    cv2.putText(img, '   left click to add a point, right click to remove it, "H" to indicate it is not visible. ', (20, 112), cv2.FONT_HERSHEY_SIMPLEX, .7, (0,0,0), 2, lineType = cv2.LINE_AA)    
-    cv2.putText(img, '   Confirm with "Y", cancel with "N".', (20, 120), cv2.FONT_HERSHEY_SIMPLEX, .5, (255,255,255), 2, lineType = cv2.LINE_AA)
+    cv2.putText(img, '   Confirm with "Y", cancel with "N".', (20, 135), cv2.FONT_HERSHEY_SIMPLEX, .7, (255,255,255), 3, lineType = cv2.LINE_AA)
-    cv2.putText(img, '   Confirm with "Y", cancel with "N".', (20, 120), cv2.FONT_HERSHEY_SIMPLEX, .5, (0,0,0), 1, lineType = cv2.LINE_AA)    
+    cv2.putText(img, '   Confirm with "Y", cancel with "N".', (20, 135), cv2.FONT_HERSHEY_SIMPLEX, .7, (0,0,0), 2, lineType = cv2.LINE_AA)    
-    cv2.putText(img, 'Use mouse wheel to zoom in and out and to pan', (20, 140), cv2.FONT_HERSHEY_SIMPLEX, .5, (255,255,255), 2, lineType = cv2.LINE_AA)
+    cv2.putText(img, 'Use mouse wheel to zoom in and out and to pan', (20, 158), cv2.FONT_HERSHEY_SIMPLEX, .7, (255,255,255), 3, lineType = cv2.LINE_AA)
-    cv2.putText(img, 'Use mouse wheel to zoom in and out and to pan', (20, 140), cv2.FONT_HERSHEY_SIMPLEX, .5, (0,0,0), 1, lineType = cv2.LINE_AA)    
+    cv2.putText(img, 'Use mouse wheel to zoom in and out and to pan', (20, 158), cv2.FONT_HERSHEY_SIMPLEX, .7, (0,0,0), 2, lineType = cv2.LINE_AA)    
     
     # Put image in a matplotlib figure for more controls
     plt.rcParams['toolbar'] = 'None'
@@ -997,15 +1010,14 @@ def recap_calibrate(ret, calib_path, calib_full_type):
         if cam != 'metadata':
             f_px = calib[cam]['matrix'][0][0]
             Dm = euclidean_distance(calib[cam]['translation'], [0,0,0])
-            print(ret[c])
             if calib_full_type=='convert_qualisys' or calib_full_type=='convert_vicon':
-                ret_m.append( np.around(ret[c]*1000, decimals=3) )
+                ret_m.append( np.around(ret[c], decimals=3) )
-                ret_px.append( np.around(ret[c] / Dm * f_px, decimals=3) )
+                ret_px.append( np.around(ret[c] / (Dm*1000) * f_px, decimals=3) )
             elif calib_full_type=='calculate_board':
                 ret_px.append( np.around(ret[c], decimals=3) )
-                ret_m.append( np.around(ret[c]*Dm / f_px, decimals=3) )
+                ret_m.append( np.around(ret[c]*Dm*1000 / f_px, decimals=3) )
 
-    logging.info(f'\n--> Residual (RMS) calibration errors for each camera are respectively {ret_px} px, \nwhich corresponds to {ret_m} m.\n')
+    logging.info(f'\n--> Residual (RMS) calibration errors for each camera are respectively {ret_px} px, \nwhich corresponds to {ret_m} mm.\n')
     logging.info(f'Calibration file is stored at {calib_path}.')
 
 

Pose2Sim/filtering.py

@@ -150,8 +150,8 @@ def kalman_filter_1d(config, col):
     - col_filtered: Filtered pandas dataframe column
     '''
 
-    trustratio = int(config.get('3d-filtering').get('kalman').get('trust_ratio'))
+    trustratio = int(config.get('filtering').get('kalman').get('trust_ratio'))
-    smooth = int(config.get('3d-filtering').get('kalman').get('smooth'))
+    smooth = int(config.get('filtering').get('kalman').get('smooth'))
     framerate = config.get('project').get('frame_rate')
     measurement_noise = 20
     process_noise = measurement_noise * trustratio
@@ -187,9 +187,9 @@ def butterworth_filter_1d(config, col):
     - col_filtered: Filtered pandas dataframe column
     '''
 
-    type = 'low' #config.get('3d-filtering').get('butterworth').get('type')
+    type = 'low' #config.get('filtering').get('butterworth').get('type')
-    order = int(config.get('3d-filtering').get('butterworth').get('order'))
+    order = int(config.get('filtering').get('butterworth').get('order'))
-    cutoff = int(config.get('3d-filtering').get('butterworth').get('cut_off_frequency'))    
+    cutoff = int(config.get('filtering').get('butterworth').get('cut_off_frequency'))    
     framerate = config.get('project').get('frame_rate')
 
     b, a = signal.butter(order/2, cutoff/(framerate/2), type, analog = False) 
@@ -223,9 +223,9 @@ def butterworth_on_speed_filter_1d(config, col):
     - col_filtered: Filtered pandas dataframe column
     '''
 
-    type = 'low' # config.get('3d-filtering').get('butterworth_on_speed').get('type')
+    type = 'low' # config.get('filtering').get('butterworth_on_speed').get('type')
-    order = int(config.get('3d-filtering').get('butterworth_on_speed').get('order'))
+    order = int(config.get('filtering').get('butterworth_on_speed').get('order'))
-    cutoff = int(config.get('3d-filtering').get('butterworth_on_speed').get('cut_off_frequency'))
+    cutoff = int(config.get('filtering').get('butterworth_on_speed').get('cut_off_frequency'))
     framerate = config.get('project').get('frame_rate')
 
     b, a = signal.butter(order/2, cutoff/(framerate/2), type, analog = False)
@@ -264,7 +264,7 @@ def gaussian_filter_1d(config, col):
     - col_filtered: Filtered pandas dataframe column
     '''
 
-    gaussian_filter_sigma_kernel = int(config.get('3d-filtering').get('gaussian').get('sigma_kernel'))
+    gaussian_filter_sigma_kernel = int(config.get('filtering').get('gaussian').get('sigma_kernel'))
 
     col_filtered = gaussian_filter1d(col, gaussian_filter_sigma_kernel)
 
@@ -284,7 +284,7 @@ def loess_filter_1d(config, col):
     - col_filtered: Filtered pandas dataframe column
     '''
 
-    kernel = config.get('3d-filtering').get('LOESS').get('nb_values_used')
+    kernel = config.get('filtering').get('LOESS').get('nb_values_used')
 
     col_filtered = col.copy()
     mask = np.isnan(col_filtered) 
@@ -313,7 +313,7 @@ def median_filter_1d(config, col):
     - col_filtered: Filtered pandas dataframe column
     '''
     
-    median_filter_kernel_size = config.get('3d-filtering').get('median').get('kernel_size')
+    median_filter_kernel_size = config.get('filtering').get('median').get('kernel_size')
     
     col_filtered = signal.medfilt(col, kernel_size=median_filter_kernel_size)
 
@@ -402,19 +402,19 @@ def recap_filter3d(config, trc_path):
     '''
 
     # Read Config
-    filter_type = config.get('3d-filtering').get('type')
+    filter_type = config.get('filtering').get('type')
-    kalman_filter_trustratio = int(config.get('3d-filtering').get('kalman').get('trust_ratio'))
+    kalman_filter_trustratio = int(config.get('filtering').get('kalman').get('trust_ratio'))
-    kalman_filter_smooth = int(config.get('3d-filtering').get('kalman').get('smooth'))
+    kalman_filter_smooth = int(config.get('filtering').get('kalman').get('smooth'))
     kalman_filter_smooth_str = 'smoother' if kalman_filter_smooth else 'filter'
-    butterworth_filter_type = 'low' # config.get('3d-filtering').get('butterworth').get('type')
+    butterworth_filter_type = 'low' # config.get('filtering').get('butterworth').get('type')
-    butterworth_filter_order = int(config.get('3d-filtering').get('butterworth').get('order'))
+    butterworth_filter_order = int(config.get('filtering').get('butterworth').get('order'))
-    butterworth_filter_cutoff = int(config.get('3d-filtering').get('butterworth').get('cut_off_frequency'))
+    butterworth_filter_cutoff = int(config.get('filtering').get('butterworth').get('cut_off_frequency'))
-    butter_speed_filter_type = 'low' # config.get('3d-filtering').get('butterworth_on_speed').get('type')
+    butter_speed_filter_type = 'low' # config.get('filtering').get('butterworth_on_speed').get('type')
-    butter_speed_filter_order = int(config.get('3d-filtering').get('butterworth_on_speed').get('order'))
+    butter_speed_filter_order = int(config.get('filtering').get('butterworth_on_speed').get('order'))
-    butter_speed_filter_cutoff = int(config.get('3d-filtering').get('butterworth_on_speed').get('cut_off_frequency'))
+    butter_speed_filter_cutoff = int(config.get('filtering').get('butterworth_on_speed').get('cut_off_frequency'))
-    gaussian_filter_sigma_kernel = int(config.get('3d-filtering').get('gaussian').get('sigma_kernel'))
+    gaussian_filter_sigma_kernel = int(config.get('filtering').get('gaussian').get('sigma_kernel'))
-    loess_filter_nb_values = config.get('3d-filtering').get('LOESS').get('nb_values_used')
+    loess_filter_nb_values = config.get('filtering').get('LOESS').get('nb_values_used')
-    median_filter_kernel_size = config.get('3d-filtering').get('median').get('kernel_size')
+    median_filter_kernel_size = config.get('filtering').get('median').get('kernel_size')
 
     # Recap
     filter_mapping_recap = {

Pose2Sim/personAssociation.py

@@ -192,7 +192,7 @@ def recap_tracking(config, error, nb_cams_excluded):
     # Read config
     project_dir = config.get('project').get('project_dir')
     if project_dir == '': project_dir = os.getcwd()
-    poseTracked_folder_name = config.get('project').get('poseTracked_folder_name')
+    poseTracked_folder_name = config.get('project').get('poseAssociated_folder_name')
     calib_folder_name = config.get('project').get('calib_folder_name')
     tracked_keypoint = config.get('personAssociation').get('tracked_keypoint')
     error_threshold_tracking = config.get('personAssociation').get('error_threshold_tracking')

Pose2Sim/triangulation.py

@@ -413,7 +413,7 @@ def triangulate_all(config):
     interp_gap_smaller_than = config.get('triangulation').get('interp_if_gap_smaller_than')
     show_interp_indices = config.get('triangulation').get('show_interp_indices')
     pose_dir = os.path.join(project_dir, pose_folder_name)
-    poseTracked_folder_name = config.get('project').get('poseTracked_folder_name')
+    poseTracked_folder_name = config.get('project').get('poseAssociated_folder_name')
     calib_dir = os.path.join(project_dir, calib_folder_name)
     calib_file = glob.glob(os.path.join(calib_dir, '*.toml'))[0]
     poseTracked_dir = os.path.join(project_dir, poseTracked_folder_name)
@@ -495,6 +495,7 @@ def triangulate_all(config):
         non_interp_frames = [[f'{seq[0]}:{seq[-1]+1}' for seq in seq_kpt if len(seq)>interp_gap_smaller_than] for seq_kpt in sequences]
     else:
         interp_frames = None
+        non_interp_frames = []
 
     # Interpolate missing values
     if interpolation_kind != 'none':

README.md

@@ -98,7 +98,7 @@ You should obtain a plot of all the 3D coordinates trajectories. You can check t
 Results are stored as .trc files in the `Demo/pose-3d` directory.
 
 *N.B.:* Default parameters have been provided in `Demo\Users\Config.toml` but can be edited.\
-**Try calibration tool by changing `calibration_type` to `calculate` instead of `convert` (more info [there](#calculate-from-scratch)).**
+*N.B.:* *Try calibration tool by changing `calibration_type` to `calculate` instead of `convert` (more info [there](#calculate-from-scratch)).*
 <br/>
 
 ## Demonstration Part-2: Obtain 3D joint angles with OpenSim  
@@ -256,6 +256,7 @@ If you already have a calibration file, set `calibration_type` type to `convert`
 
 - **With a board:**
   > *N.B.:* Try the calibration tool on the Demo by changing `calibration_type` to `calculate` instead of `convert` in `Config.toml`.\
+  For the sake of practicality, there are voluntarily few images for intrinsics, and few clicked points for extrinsics. *You should use more of them.* In spite of this, your reprojection error should be under 1-2 cm, which [does not hinder the quality of kinematic results in practice](https://www.mdpi.com/1424-8220/21/19/6530/htm).
   
   - **Calculate intrinsic parameters:**