add example in iris demo to load video from video file instead of webcam

face_detection.py

@@ -1,6 +1,6 @@
 import mediapipe as mp
 
-from webcam import WebcamSource
+from videosource import WebcamSource
 
 mp_drawing = mp.solutions.drawing_utils
 mp_face_detection = mp.solutions.face_detection

face_mesh.py

@@ -1,6 +1,6 @@
 import mediapipe as mp
 
-from webcam import WebcamSource
+from videosource import WebcamSource
 
 mp_drawing = mp.solutions.drawing_utils
 mp_face_mesh = mp.solutions.face_mesh

hands.py

@@ -1,6 +1,6 @@
 import mediapipe as mp
 
-from webcam import WebcamSource
+from videosource import WebcamSource
 
 mp_drawing = mp.solutions.drawing_utils
 mp_hands = mp.solutions.hands

head_posture.py

@@ -2,7 +2,7 @@
 import mediapipe as mp
 import numpy as np
 
-from webcam import WebcamSource
+from videosource import WebcamSource
 
 from custom.face_geometry import (  # isort:skip
     PCF,

holistic.py

@@ -1,6 +1,6 @@
 import mediapipe as mp
 
-from webcam import WebcamSource
+from videosource import WebcamSource
 
 mp_drawing = mp.solutions.drawing_utils
 mp_holistic = mp.solutions.holistic

iris.py

@@ -1,45 +1,21 @@
+import argparse
+
 import cv2
 import mediapipe as mp
 import numpy as np
 
-from webcam import WebcamSource
-
-from custom.iris_lm_depth import (  # isort:skip
-    calculate_iris_depth,
-    detect_iris,
-    from_landmarks_to_depth,
-)
-
-from custom.core import (  # isort:skip
-    detections_to_rect,
-    landmarks_to_detections,
-    slice_from_roi,
-    transform_rect,
-)
+from custom.iris_lm_depth import from_landmarks_to_depth
+from videosource import FileSource, WebcamSource
 
 mp_face_mesh = mp.solutions.face_mesh
 
 points_idx = [33, 133, 362, 263, 61, 291, 199]
 points_idx = list(set(points_idx))
 points_idx.sort()
 
-frame_height, frame_width, channels = (720, 1280, 3)
-image_size = (frame_width, frame_height)
-
 left_eye_landmarks_id = np.array([33, 133])
 right_eye_landmarks_id = np.array([362, 263])
 
-frame_height, frame_width, channels = (720, 1280, 3)
-frame_size = np.array((frame_width, frame_height))
-
-# pseudo camera internals
-focal_length = frame_width
-center = (frame_width / 2, frame_height / 2)
-camera_matrix = np.array(
-    [[focal_length, 0, center[0]], [0, focal_length, center[1]], [0, 0, 1]],
-    dtype="double",
-)
-
 dist_coeff = np.zeros((4, 1))
 
 YELLOW = (0, 255, 255)
@@ -50,21 +26,28 @@
 LARGE_CIRCLE_SIZE = 2
 
 
-def main():
-    source = WebcamSource(width=frame_width, height=frame_height)
+def main(inp):
+    if inp is None:
+        frame_height, frame_width = (720, 1280)
+        source = WebcamSource(width=frame_width, height=frame_height)
+    else:
+        source = FileSource(inp)
+        frame_width, frame_height = (int(i) for i in source.get_image_size())
+
+    image_size = (frame_width, frame_height)
+
+    # pseudo camera internals
+    focal_length = frame_width
 
     landmarks = None
     smooth_left_depth = -1
     smooth_right_depth = -1
     smooth_factor = 0.1
-    face_mesh = mp_face_mesh.FaceMesh(
+
+    with mp_face_mesh.FaceMesh(
         static_image_mode=False,
         min_detection_confidence=0.5,
         min_tracking_confidence=0.5,
-    )
-
-    with mp_face_mesh.FaceMesh(
-        min_detection_confidence=0.5, min_tracking_confidence=0.5
     ) as face_mesh:
 
         for idx, (frame, frame_rgb) in enumerate(source):
@@ -129,7 +112,7 @@ def main():
 
                 # draw subset of facemesh
                 for ii in points_idx:
-                    pos = (frame_size * landmarks[:2, ii]).astype(np.int32)
+                    pos = (np.array(image_size) * landmarks[:2, ii]).astype(np.int32)
                     frame = cv2.circle(frame, tuple(pos), LARGE_CIRCLE_SIZE, GREEN, -1)
 
                 # draw eye contours
@@ -140,7 +123,7 @@ def main():
                     ]
                 )
                 for landmark in eye_landmarks:
-                    pos = (frame_size * landmark[:2]).astype(np.int32)
+                    pos = (np.array(image_size) * landmark[:2]).astype(np.int32)
                     frame = cv2.circle(frame, tuple(pos), SMALL_CIRCLE_SIZE, RED, -1)
 
                 # draw iris landmarks
@@ -151,7 +134,7 @@ def main():
                     ]
                 )
                 for landmark in iris_landmarks:
-                    pos = (frame_size * landmark[:2]).astype(np.int32)
+                    pos = (np.array(image_size) * landmark[:2]).astype(np.int32)
                     frame = cv2.circle(frame, tuple(pos), SMALL_CIRCLE_SIZE, YELLOW, -1)
 
                 # write depth values into frame
@@ -173,4 +156,10 @@ def main():
 
 
 if __name__ == "__main__":
-    main()
+    parser = argparse.ArgumentParser(description="Choose video file otherwise webcam is used.")
+    parser.add_argument(
+        "-i", metavar="path-to-file", type=str, help="Path to video file"
+    )
+
+    args = parser.parse_args()
+    main(args.i)

objectron.py

@@ -1,6 +1,6 @@
 import mediapipe as mp
 
-from webcam import WebcamSource
+from videosource import WebcamSource
 
 mp_drawing = mp.solutions.drawing_utils
 mp_objectron = mp.solutions.objectron

pose.py

@@ -1,7 +1,7 @@
 import mediapipe as mp
 import numpy as np
 
-from webcam import WebcamSource
+from videosource import WebcamSource
 
 mp_drawing = mp.solutions.drawing_utils
 mp_pose = mp.solutions.pose

selfie_segmentation.py

@@ -1,8 +1,7 @@
 import cv2
 import mediapipe as mp
 import numpy as np
-
-from webcam import WebcamSource
+from videosource import WebcamSource
 
 mp_drawing = mp.solutions.drawing_utils
 mp_selfie_segmentation = mp.solutions.selfie_segmentation

videosource.py

@@ -0,0 +1,80 @@
+import cv2
+import numpy as np
+from PIL import Image
+
+
+class VideoSource:
+    def __init__(self, display=False, dtype=np.uint8):
+        self._name = "VideoSource"
+        self._capture = None
+        self._display = display
+        self._dtype = dtype
+
+    def get_fps(self):
+        return self._capture.get(cv2.CAP_PROP_FPS)
+
+    def get_frame_count(self):
+        return int(self._capture.get(cv2.CAP_PROP_FRAME_COUNT))
+
+    def get_image_size(self):
+        width = self._capture.get(cv2.CAP_PROP_FRAME_WIDTH)
+        height = self._capture.get(cv2.CAP_PROP_FRAME_HEIGHT)
+        return width, height
+
+    def release(self):
+        self._capture.release()
+        cv2.destroyAllWindows()
+
+    def __iter__(self):
+        self._capture.isOpened()
+        return self
+
+    def __next__(self):
+        ret, frame = self._capture.read()
+
+        if self._display:
+            cv2.imshow(f"{self._name} - FPS: {self.get_fps()}", frame)
+
+        if not ret:
+            raise StopIteration
+
+        if cv2.waitKey(1) & 0xFF == ord("q"):
+            raise StopIteration
+
+        cv2_im_rgb = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
+        frame_rgb = np.asarray(Image.fromarray(cv2_im_rgb), dtype=self._dtype)
+
+        return frame, frame_rgb
+
+    def __del__(self):
+        self.release()
+
+    def show(self, frame):
+        cv2.imshow(f"{self._name} - FPS: {self.get_fps()}", frame)
+
+
+class WebcamSource(VideoSource):
+    def __init__(
+        self,
+        camera_id=0,
+        width=1280,
+        height=720,
+        fps=30,
+        autofocus=0,
+        absolute_focus=75,
+        display=False,
+    ):
+        super().__init__(display)
+        self._capture = cv2.VideoCapture(camera_id)
+        self._capture.set(cv2.CAP_PROP_FRAME_WIDTH, width)
+        self._capture.set(cv2.CAP_PROP_FRAME_HEIGHT, height)
+        self._capture.set(cv2.CAP_PROP_FOURCC, cv2.VideoWriter_fourcc(*"MJPG"))
+        self._capture.set(cv2.CAP_PROP_FPS, fps)
+        self._capture.set(cv2.CAP_PROP_AUTOFOCUS, autofocus)
+        self._capture.set(cv2.CAP_PROP_FOCUS, absolute_focus / 255)
+
+
+class FileSource(VideoSource):
+    def __init__(self, file_path, display=False):
+        super().__init__(display)
+        self._capture = cv2.VideoCapture(str(file_path))