adjustments

example.py

@@ -2,20 +2,54 @@
 Demonstration of the GazeTracking library.
 Check the README.md for complete documentation.
 """
-
+import datetime
+import os
 import cv2
 from gaze_tracking import GazeTracking
+import time
+import utils
+import numpy as np
+
+stim_dir = "left"
+stim_dist = "23"
+stim_trial = "XX"
+stim_time = datetime.datetime.now().strftime("%Y-%m-%d_%H-%M-%S")
+savefile = f"{stim_dir}_{stim_dist}_{stim_trial}_{stim_time}.npy"
+
+print_interval = 1000 # n frames to update to console
+new_width = 400
+new_height = 300
+
+center = (new_width, new_height)
+x = center[1] / 2 - new_width / 2
+y = center[0] / 2 - new_height / 2
 
 gaze = GazeTracking()
-webcam = cv2.VideoCapture(0)
+webcam = cv2.VideoCapture(max(utils.list_webcams()))
+init_time = time.perf_counter()
+framecount = 0
+
+# TODO: get monitor resolution, adjust params
+monitor_res_y = 1080
+monitor_res_x = 720
+patient_distance = 40 # cm
+
+study_eye = "left"
+study_eye_positions = []
 
 while True:
+    loop_start_time = time.perf_counter()
+    framecount += 1
+
     # We get a new frame from the webcam
     _, frame = webcam.read()
 
-    # We send this frame to GazeTracking to analyze it
-    gaze.refresh(frame)
+    # Resize the image
+    frame = cv2.resize(frame, (new_width, new_height))
+    crop_img = frame[int(y):int(y + new_height), int(x):int(x + new_width)]
 
+    # We send this frame to GazeTracking to analyze it
+    gaze.refresh(frame) # this is the most expensive operation
     frame = gaze.annotated_frame()
     text = ""
 
@@ -32,13 +66,33 @@
 
     left_pupil = gaze.pupil_left_coords()
     right_pupil = gaze.pupil_right_coords()
+    if left_pupil is None and right_pupil is None:
+        print ("No pupils detected")
+    if study_eye == "left":
+        if left_pupil is not None:
+            study_eye_positions.append(left_pupil)
+            deviation = utils.distance_from_fixation((crop_img.shape[0] // 2, crop_img.shape[1] // 2), left_pupil)
+            # study_eye_deviations.append(deviation)
+    elif study_eye == "right":
+        if right_pupil is not None:
+            study_eye_positions.append(right_pupil)
+            deviation = utils.distance_from_fixation((crop_img.shape[0] // 2, crop_img.shape[1] // 2), right_pupil)
+            # study_eye_deviations.append(deviation)
     cv2.putText(frame, "Left pupil:  " + str(left_pupil), (90, 130), cv2.FONT_HERSHEY_DUPLEX, 0.9, (147, 58, 31), 1)
     cv2.putText(frame, "Right pupil: " + str(right_pupil), (90, 165), cv2.FONT_HERSHEY_DUPLEX, 0.9, (147, 58, 31), 1)
 
     cv2.imshow("Demo", frame)
 
     if cv2.waitKey(1) == 27:
         break
-
+    fps = 1/(time.perf_counter() - loop_start_time)
+    if framecount % print_interval == 0:
+        print(f"FPS: {fps}")
+
+study_eye_deviations = [utils.distance_from_fixation((crop_img.shape[0] // 2, crop_img.shape[1] // 2), pos) for pos in study_eye_positions]
+print ("Average deviation of study eye from fixation: ", np.round(np.mean(study_eye_deviations), 4), "over ", framecount, "frames")
+print ("Std dev of deviation of study eye from fixation: ", np.round(np.std(study_eye_deviations), 4), "over ", framecount, "frames")
+print (os.getcwd())
+np.save(os.path.join("data", savefile), study_eye_positions)
 webcam.release()
 cv2.destroyAllWindows()

utils.py

@@ -0,0 +1,20 @@
+
+
+import cv2
+
+def list_webcams():
+    index = 0
+    arr = []
+    while True:
+        cap = cv2.VideoCapture(index)
+        if not cap.read()[0]:
+            break
+        else:
+            arr.append(index)
+        cap.release()
+        index += 1
+    return arr
+
+
+def distance_from_fixation(origin, point):
+    return ((origin[0] - point[0]) ** 2 + (origin[1] - point[1]) ** 2) ** 0.5