mirror of
https://github.com/EyeTrackVR/EyeTrackVR-Docs.git
synced 2025-09-26 23:09:15 +08:00
173 lines
5.7 KiB
Python
173 lines
5.7 KiB
Python
import cv2
|
|
import numpy as np
|
|
import threading
|
|
import time
|
|
import os
|
|
|
|
try:
|
|
fy= open("camport.txt","r+")
|
|
camport = fy.read().strip()
|
|
fy.close
|
|
except:
|
|
camport = 0
|
|
cam = open('camport.txt', 'w+')
|
|
cam.write('0')
|
|
cam.close
|
|
print('Error: Run Gui first and adjust camera port')
|
|
time.sleep(1)
|
|
bl = False
|
|
cap = cv2.VideoCapture(int(camport))
|
|
width = cap.get(cv2.CAP_PROP_FRAME_WIDTH)
|
|
height = cap.get(cv2.CAP_PROP_FRAME_HEIGHT)
|
|
print(width, height)
|
|
|
|
|
|
def lefteye():
|
|
x = 1 #defining variables for smoothing
|
|
y = 1
|
|
x1 = 1
|
|
y1 = 1
|
|
h = 1 #defines hight value so blink check doesnt error out
|
|
|
|
def smooth(x, x1, y, y1):
|
|
xsmooth = (x1 + x) / 2
|
|
ysmooth = (y1 + y) / 2
|
|
print('x filtered:', xsmooth, 'y filtered:', ysmooth, 'x raw:', x, 'y raw:', y)
|
|
#print('x raw:', x, 'y raw:', y)
|
|
|
|
|
|
while True: #loop for eye detection
|
|
ret, frame = cap.read()
|
|
if ret is False:
|
|
break
|
|
|
|
################################################ Reads values set in gui, needs rework to single config file
|
|
try:
|
|
fy= open("valueY.txt","r+")
|
|
vy = fy.read().strip()
|
|
fy.close
|
|
except:
|
|
vy = str(height)
|
|
v= open("valueY.txt","w+")
|
|
v.write(vy)
|
|
v.close()
|
|
print('WARNING: Run Gui first and adjust Value Y')
|
|
#time.sleep(2)
|
|
################################################
|
|
try:
|
|
fx= open("valueX.txt","r+")
|
|
vx = fx.read().strip()
|
|
fx.close
|
|
except:
|
|
vx = str(width)
|
|
fx= open("valueX.txt","w+")
|
|
fx.write(vx)
|
|
fx.close()
|
|
print('WARNING: Run Gui first and adjust Value X')
|
|
#time.sleep(2)
|
|
################################################
|
|
try:
|
|
vyll= open("valueYl.txt","r+")
|
|
vyl = vyll.read().strip()
|
|
vyll.close
|
|
except:
|
|
vyll= open("valueYl.txt","w+")
|
|
vyll.write('1')
|
|
vyll.close
|
|
vyl = 1
|
|
print('WARNING: Run Gui first and adjust Value Y L')
|
|
time.sleep(1)
|
|
################################################
|
|
try:
|
|
vxll= open("valueXl.txt","r+")
|
|
vxl = vxll.read().strip()
|
|
vxll.close
|
|
except:
|
|
vxl = 1
|
|
vyll= open("valueXl.txt","w+")
|
|
vyl = vyll.write('1')
|
|
vyll.close
|
|
print('Error: Run Gui first and adjust Value X L')
|
|
time.sleep(1)
|
|
################################################
|
|
try:
|
|
thresh= open("thresh.txt","r+")
|
|
threshr = thresh.read().strip()
|
|
thresh.close
|
|
except:
|
|
thresh= open("thresh.txt","w+")
|
|
threshr = thresh.write('19')
|
|
thresh.close
|
|
print('WARNING: Run Gui first and adjust threshold value')
|
|
threshr = 19
|
|
time.sleep(1)
|
|
|
|
|
|
################################################
|
|
# try:
|
|
try: # trys at set size if it errors it will revert to working size/ doesnt do what was orrigionally planed, it kinda helps
|
|
roi = frame[int(vxl): int(float(vy)), int(vyl): int(float(vx))]
|
|
except:
|
|
roi = frame[100: 300, 200: 316]
|
|
try:
|
|
x1 = x
|
|
y1 = y
|
|
rows, cols, _ = roi.shape
|
|
gray_roi = cv2.cvtColor(roi, cv2.COLOR_BGR2GRAY)
|
|
gray_roi = cv2.GaussianBlur(gray_roi, (7, 7), 0)
|
|
|
|
_, threshold = cv2.threshold(gray_roi, int(threshr), 255, cv2.THRESH_BINARY_INV)
|
|
contours, _ = cv2.findContours(threshold, cv2.RETR_TREE, cv2.CHAIN_APPROX_SIMPLE)
|
|
contours = sorted(contours, key=lambda x: cv2.contourArea(x), reverse=True)
|
|
|
|
for cnt in contours:
|
|
|
|
(x, y, w, h) = cv2.boundingRect(cnt)
|
|
|
|
|
|
openes = 95
|
|
if h <= 25: #way to find if eye is closed and sets value (hopefully will train tensorflow model for correct openess detection)
|
|
openes = 0
|
|
|
|
smooth(x, x1, y, y1)
|
|
#print('Left: x:', x, 'y:', y, 'openess:', openes)
|
|
#print('Right: x:', x, 'y:', y, 'openess:', openes)
|
|
|
|
cv2.line(threshold, (x + int(w/2), 0), (x + int(w/2), rows), (255, 0, 0), 1) #visualizes eyetracking on threshold
|
|
cv2.line(threshold, (0, y + int(h/2)), (cols, y + int(h/2)), (255, 0, 0), 1)
|
|
cv2.drawContours(threshold, [cnt], -1, (255, 0, 0), 3)
|
|
cv2.rectangle(threshold, (x, y), (x + w, y + h), (255, 0, 0), 2)
|
|
|
|
cv2.line(gray_roi, (x + int(w/2), 0), (x + int(w/2), rows), (255, 0, 0), 1) #visualizes eyetracking on greyscale
|
|
cv2.line(gray_roi, (0, y + int(h/2)), (cols, y + int(h/2)), (255, 0, 0), 1)
|
|
cv2.drawContours(gray_roi, [cnt], -1, (255, 0, 0), 3)
|
|
cv2.rectangle(gray_roi, (x, y), (x + w, y + h), (255, 0, 0), 2)
|
|
break
|
|
|
|
if h == 0: #basically useless lol
|
|
print('no eye detected"')
|
|
|
|
|
|
|
|
cv2.imshow("Threshold", threshold)
|
|
cv2.imshow("GreyScale", gray_roi)
|
|
#cv2.imshow("Roi", roi)
|
|
key = cv2.waitKey(30)
|
|
except:
|
|
print('ERROR 1: Something went wrong trying to track your eye.')
|
|
|
|
|
|
|
|
print('###############> initailizing <###############')
|
|
|
|
cv2.destroyAllWindows()
|
|
if __name__ == '__main__':
|
|
lefteye()
|
|
#t1 = threading.Thread(target=lefteye)
|
|
#t2 = threading.Thread(target=righteye)
|
|
#t1.start()
|
|
#t2.start()
|
|
#t1.join()
|
|
#t2.join()
|
|
print('Running:')
|