diff --git a/ar_tag_pose/detector_board_main.py b/ar_tag_pose/detector_board_main.py
index 6dd95b5..2a074dc 100644
--- a/ar_tag_pose/detector_board_main.py
+++ b/ar_tag_pose/detector_board_main.py
@@ -100,6 +100,7 @@ class Detector(abc.ABC):
 
 		frame_go = True
 		frame_num = 0
+		out = None
 		while cap.isOpened():
 			if self.src_images:
 				cap.set(cv.CAP_PROP_POS_FRAMES, frame_num)
@@ -151,12 +152,31 @@ class Detector(abc.ABC):
 
 			img_scaled = cv.resize(img, dsize=None, fx=0.5, fy=0.5)
 			cv.imshow('img', img_scaled)
+			out = out_writer(out, img_scaled)
 
 			if frame_go:
 				frame_num += 1
 
+		cap.release()
+		out_writer(out, None)
+
 		cv.destroyAllWindows()
 
+def out_writer(writer: cv.VideoWriter|None, frame: np.ndarray|None) -> cv.VideoWriter:
+	if writer is None:
+		fourcc = cv.VideoWriter.fourcc(*"mp4v")
+		frame_size = (frame.shape[1], frame.shape[0])
+		writer = cv.VideoWriter(filename=f"{RESULTS_FOLDER}/out.mp4", fourcc=fourcc, fps=30.0, frameSize=frame_size)
+
+	if writer is not None:
+		if frame is not None:
+			writer.write(frame)
+		else:
+			writer.release()
+			writer = None
+
+	return writer
+
 
 def main():
 	# construct the argument parser and parse the arguments