deeplabcut.utils.auxiliaryfunctions_3d

DeepLabCut2.0 Toolbox (deeplabcut.org) © A. & M. Mathis Labs https://github.com/DeepLabCut/DeepLabCut Please see AUTHORS for contributors.

https://github.com/DeepLabCut/DeepLabCut/blob/master/AUTHORS Licensed under GNU Lesser General Public License v3.0

Functions:

Name	Description
Foldernames3Dproject	Definitions of subfolders in 3D projects.
Get_list_of_triangulated_and_videoFiles	Returns the list of triangulated h5 and the corresponding video files.
compute_triangulation_calibration_images	Performs triangulation of the calibration images.
cross_view_match_dataframes	Computes the costs and matched voting for tracks between a camera pair.
get_camerawise_videos	This function returns the list of videos corresponding to the camera names

Foldernames3Dproject

Foldernames3Dproject(cfg_3d)

Definitions of subfolders in 3D projects.

Source code in deeplabcut/utils/auxiliaryfunctions_3d.py

def Foldernames3Dproject(cfg_3d):
    """Definitions of subfolders in 3D projects."""

    img_path = os.path.join(cfg_3d["project_path"], "calibration_images")
    path_corners = os.path.join(cfg_3d["project_path"], "corners")
    path_camera_matrix = os.path.join(cfg_3d["project_path"], "camera_matrix")
    path_undistort = os.path.join(cfg_3d["project_path"], "undistortion")
    path_removed_images = os.path.join(cfg_3d["project_path"], "removed_calibration_images")

    return (
        img_path,
        path_corners,
        path_camera_matrix,
        path_undistort,
        path_removed_images,
    )

Get_list_of_triangulated_and_videoFiles

Get_list_of_triangulated_and_videoFiles(filepath, videotype, scorer_3d, cam_names, videofolder)

Returns the list of triangulated h5 and the corresponding video files.

Source code in deeplabcut/utils/auxiliaryfunctions_3d.py

def Get_list_of_triangulated_and_videoFiles(filepath, videotype, scorer_3d, cam_names, videofolder):
    """Returns the list of triangulated h5 and the corresponding video files."""

    prefix = []
    suffix = []
    file_list = []
    string_to_search = scorer_3d + ".h5"

    # Checks if filepath is a directory
    if [os.path.isdir(i) for i in filepath] == [True]:
        """Analyzes all the videos in the directory."""
        print("Analyzing all the videos in the directory")
        videofolder = filepath[0]
        cwd = os.getcwd()
        os.chdir(videofolder)
        triangulated_file_list = [fn for fn in os.listdir(os.curdir) if (string_to_search in fn)]
        video_list = get_camerawise_videos(videofolder, cam_names, videotype)
        os.chdir(cwd)
        triangulated_folder = videofolder
    else:
        triangulated_file_list = [str(Path(fn).name) for fn in filepath if (string_to_search in fn)]
        triangulated_folder = [str(Path(fn).parents[0]) for fn in filepath if (string_to_search in fn)]
        triangulated_folder = triangulated_folder[0]

        if videofolder is None:
            videofolder = str(Path(filepath[0]).parents[0])
        video_list = get_camerawise_videos(videofolder, cam_names, videotype)

    # Get the filename of the triangulated file excluding the scorer name and remove any '-' or _ from it
    filename = [i.split(string_to_search)[0] for i in triangulated_file_list]
    for i in range(len(filename)):
        if filename[i][-1] == "_" or filename[i][-1] == "-":
            filename[i] = filename[i][:-1]
        if filename[i][0] == "_" or filename[i][0] == "-":
            filename[i] = filename[i][1:]

    # Get the suffix and prefix of the video filenames so that they can be
    # used for matching the triangulated file names.
    for i in range(len(video_list)):
        pre = [
            str(Path(video_list[i][0]).stem).split(cam_names[0])[0],
            str(Path(video_list[i][1]).stem).split(cam_names[1])[0],
        ]
        suf = [
            str(Path(video_list[i][0]).stem).split(cam_names[0])[-1],
            str(Path(video_list[i][1]).stem).split(cam_names[1])[-1],
        ]
        for i in range(len(cam_names)):
            if pre[i] == "":
                pass
            elif pre[i][-1] == "_" or pre[i][-1] == "-":
                pre[i] = pre[i][:-1]
            if suf[i] == "":
                pass
            elif suf[i][0] == "_" or suf[i][0] == "-":
                suf[i] = suf[i][1:]
        suffix.append(suf)
        prefix.append(pre)

    # Match the suffix and prefix with the triangulated file name and return
    # the list with triangulated file and corresponding video files.
    for k in range(len(filename)):
        for j in range(len(prefix)):
            if (prefix[j][0] in filename[k] and prefix[j][1] in filename[k]) and (
                suffix[j][0] in filename[k] and suffix[j][1] in filename[k]
            ):
                triangulated_file = glob.glob(
                    os.path.join(
                        triangulated_folder,
                        str("*" + filename[k] + "*" + string_to_search),
                    )
                )
                vfiles = get_camerawise_videos(videofolder, cam_names, videotype)
                vfiles = [z for z in vfiles if prefix[j][0] in z[0] and suffix[j][0] in z[1]][0]
                file_list.append(triangulated_file + vfiles)

    return file_list

compute_triangulation_calibration_images

compute_triangulation_calibration_images(
    stereo_matrix, projectedPoints1, projectedPoints2, path_undistort, cfg_3d, plot=True
)

Performs triangulation of the calibration images.

Source code in deeplabcut/utils/auxiliaryfunctions_3d.py

def compute_triangulation_calibration_images(
    stereo_matrix, projectedPoints1, projectedPoints2, path_undistort, cfg_3d, plot=True
):
    """Performs triangulation of the calibration images."""
    triangulate = []
    P1 = stereo_matrix["P1"]
    P2 = stereo_matrix["P2"]
    cmap = cfg_3d["colormap"]
    colormap = plt.get_cmap(cmap)
    markerSize = cfg_3d["dotsize"]
    markerType = cfg_3d["markerType"]

    for i in range(projectedPoints1.shape[0]):
        X_l = triangulatePoints(P1, P2, projectedPoints1[i], projectedPoints2[i])
        triangulate.append(X_l)
    triangulate = np.asanyarray(triangulate)

    # Plotting
    if plot:
        col = colormap(np.linspace(0, 1, triangulate.shape[0]))
        fig = plt.figure()
        ax = fig.add_subplot(111, projection="3d")

        for i in range(triangulate.shape[0]):
            xs = triangulate[i, 0, :]
            ys = triangulate[i, 1, :]
            zs = triangulate[i, 2, :]
            ax.scatter(xs, ys, zs, c=col[i], marker=markerType, s=markerSize)
            ax.set_xlabel("X")
            ax.set_ylabel("Y")
            ax.set_zlabel("Z")
        plt.savefig(os.path.join(str(path_undistort), "checkerboard_3d.png"))
    return triangulate

cross_view_match_dataframes

cross_view_match_dataframes(df1, df2, F)

Computes the costs and matched voting for tracks between a camera pair.

df: Data read from .h5 track file F: fundamental matrix from OpenCV

Source code in deeplabcut/utils/auxiliaryfunctions_3d.py

def cross_view_match_dataframes(df1, df2, F):
    """Computes the costs and matched voting for tracks between a camera pair.

    df: Data read from .h5 track file
    F: fundamental matrix from OpenCV
    """

    tracks1 = _reconstruct_tracks_as_tracklets(df1)
    tracks2 = _reconstruct_tracks_as_tracklets(df2)
    costs, voting = _associate_paired_view_tracks(tracks1, tracks2, F)

    return costs, voting

get_camerawise_videos

get_camerawise_videos(path, cam_names, videotype)

This function returns the list of videos corresponding to the camera names specified in the cam_names. e.g. if cam_names = ['camera-1','camera-2']

then it will return [['somename-camera-1-othername.avi', 'somename- camera-2-othername.avi']]

Source code in deeplabcut/utils/auxiliaryfunctions_3d.py

def get_camerawise_videos(path, cam_names, videotype):
    """This function returns the list of videos corresponding to the camera names
    specified in the cam_names. e.g. if cam_names = ['camera-1','camera-2']

    then it will return [['somename-camera-1-othername.avi', 'somename-
    camera-2-othername.avi']]
    """
    import glob
    from pathlib import Path

    vid = []

    # Find videos only specific to the cam names
    videos = [glob.glob(os.path.join(path, str("*" + cam_names[i] + "*" + videotype))) for i in range(len(cam_names))]
    videos = [y for x in videos for y in x]

    # Exclude the labeled video files
    if "." in videotype:
        file_to_exclude = str("labeled" + videotype)
    else:
        file_to_exclude = str("labeled." + videotype)
    videos = [v for v in videos if os.path.isfile(v) and file_to_exclude not in v]
    video_list = []
    cam = cam_names[0]  # camera1
    vid.append(
        [name for name in glob.glob(os.path.join(path, str("*" + cam + "*" + videotype)))]
    )  # all videos with cam
    # print("here is what I found",vid)
    for k in range(len(vid[0])):
        if cam in str(Path(vid[0][k]).stem):
            ending = Path(vid[0][k]).suffix
            pref = str(Path(vid[0][k]).stem).split(cam)[0]
            suf = str(Path(vid[0][k]).stem).split(cam)[1]
            if pref == "":
                if suf == "":
                    print("Strange naming convention on your part. Respect.")
                else:
                    putativecam2name = os.path.join(path, cam_names[1] + suf + ending)
            else:
                if suf == "":
                    putativecam2name = os.path.join(path, pref + cam_names[1] + ending)
                else:
                    putativecam2name = os.path.join(path, pref + cam_names[1] + suf + ending)
            # print([os.path.join(path,pref+cam+suf+ending),putativecam2name])
            if os.path.isfile(putativecam2name):
                # found a pair!!!
                video_list.append([os.path.join(path, pref + cam + suf + ending), putativecam2name])
    return video_list