Tutorial¶

With eyetracker-ng it is very easy to implement custom image processing algorithms. All you have to do is include your code into the ‘your code’ section in the following examples.

import eyetracker

cam_id = eyetracker.camera.camera.lookForCameras()

cam = eyetracker.camera.camera.Camera(cam_id['Camera_1'])

while True:
    frame = cam.frame()
    
    ''' your code '''

    ''' end of your code'''

    ''' press q or ESC to quit'''

    key = eyetracker.camera.display.displayImage(frame)
    if key == 27 or key == ord('q'):
        break

import eyetracker.camera.camera as camera
import eyetracker.camera.display as display
from eyetracker.analysis.detect import glint, pupil
from eyetracker.analysis.processing import gray2bgr, bgr2gray, mark, threshold, thresholds
import cv2
import numpy as np

def nothing(x):
    pass

cam_id = camera.lookForCameras()
cams = cam_id.keys()
cam = camera.Camera(cam_id['Camera_1'])
cam_nr = 0
window = 'decide on the name'

cv2.namedWindow(window)
cv2.namedWindow('choose your camera')
cv2.createTrackbar('threshold value', window, 0, 255, nothing)
cv2.createTrackbar('threshold type', window, 0, 5, nothing)

key = 0

while key != ord('y'):
    frame = cam.frame()
    key = display.displayImage(frame, 'choose your camera')
    if key == ord('>'):
        cam_nr += 1
        cam.close()
        cam = camera.Camera(cam_id[cams[cam_nr % len(cams)]])

cv2.destroyWindow('choose your camera')

threshold_types = thresholds.keys()

print 'Indexes corresponding to threshold types:'
for i, v in enumerate(threshold_types):
    print i, ':', v

while True:
    thresh_v = cv2.getTrackbarPos('threshold value', window)
    thresh_t = int(cv2.getTrackbarPos('threshold type', window))

    frame = cam.frame()

    frame_gray = bgr2gray(frame)

    frame_thresh = threshold(frame_gray, thresh_v, 
                             thresh_type=threshold_types[thresh_t])

    where_glint = glint(frame_gray)

    where_pupil = pupil(frame_thresh)
    
    ''' your code '''

    mark(frame, where_glint)

    mark(frame, where_pupil, color='blue')
    
    frame_bgr = gray2bgr(frame_thresh)

    new_image = np.hstack((frame, frame_bgr))

    ''' end of your code'''

    ''' press q or ESC to quit'''

    key = display.displayImage(new_image, where=window)
    if key == 27 or key == ord('q'):
        break

eyetracker-ng 0.1-1 documentation

Tutorial

Tutorial¶