Prequisites to learn and run the below code
I started learning pyton and opencv
These are the list of urls i followed to learn
http://docs.opencv.org/trunk/
https://www.youtube.com/
These are the list of urls i followed to learn
http://docs.opencv.org/trunk/
https://www.youtube.com/
pip
easy_install
http://scikit-learn.org/
http://scikit-image.org/docs/
and wrote the below code by using the code in the above urls
import cv2
from skimage.feature import local_binary_pattern
import numpy as np
face_cascade = cv2.CascadeClassifier('haarcascade_frontalface_default.xml')
eye_cascade = cv2.CascadeClassifier('rightEye.xml')
nose_cascade = cv2.CascadeClassifier('haarcascade_mcs_nose.xml')
METHOD = 'uniform'
radius = 2
n_points = 8 * radius
def hist(ax, lbp):
n_bins = lbp.max() + 1
return ax.hist(lbp.ravel(), normed=True, bins=n_bins, range=(0, n_bins),
facecolor='0.5')
def kullback_leibler_divergence(p, q):
p = np.asarray(p)
q = np.asarray(q)
filt = np.logical_and(p != 0, q != 0)
return np.sum(p[filt] * np.log2(p[filt] / q[filt]))
def match(refs, img):
best_score = 10
best_name = None
lbp = local_binary_pattern(img, n_points, radius, METHOD)
n_bins = lbp.max() + 1
hist, _ = np.histogram(lbp, normed=True, bins=n_bins, range=(0, n_bins))
for name, ref in refs.items():
ref_hist, _ = np.histogram(ref, normed=True, bins=n_bins,
range=(0, n_bins))
score = kullback_leibler_divergence(hist, ref_hist)
if score < best_score:
best_score = score
best_name = name
return best_name
brick = cv2.imread('eclosed.jpg')
brick= cv2.cvtColor(brick, cv2.COLOR_BGR2GRAY)
brick= cv2.resize(brick, (21,21),cv2.INTER_AREA)
grass = cv2.imread('eopen1.jpg')
grass = cv2.cvtColor(grass, cv2.COLOR_BGR2GRAY)
grass= cv2.resize(grass, (21,21),cv2.INTER_AREA)
wall = cv2.imread('eye2.jpg')
wall=cv2.cvtColor(wall, cv2.COLOR_BGR2GRAY)
wall= cv2.resize(wall, (21,21),cv2.INTER_AREA)
refs = {
'brick': local_binary_pattern(brick, n_points, radius, METHOD),
'grass': local_binary_pattern(grass, n_points, radius, METHOD),
'wall': local_binary_pattern(wall, n_points, radius, METHOD)
}
cap = cv2.VideoCapture('sample1.avi')
if __name__ == "__main__":
while(cap.isOpened()):
ret, img = cap.read()
#img = cv2.imread('closed.png')
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
faces = face_cascade.detectMultiScale(gray, 1.3, 5)
for (x,y,w,h) in faces:
img = cv2.rectangle(img,(x,y),(x+w,y+h),(255,0,0),2)
roi_gray = gray[y:y+h, x:x+w]
roi_color = img[y:y+h, x:x+w]
eyes = eye_cascade.detectMultiScale(roi_gray)
for (ex,ey,ew,eh) in eyes:
cv2.rectangle(roi_color,(ex,ey),(ex+ew,ey+eh),(0,255,0),2)
#le = roi_color[eyes[0,1]:eyes[0,1]+eyes[0,3], eyes[0,0]:eyes[0,0]+eyes[0,2]]
re = roi_color[eyes[1,1]:eyes[1,1]+eyes[1,3], eyes[1,0]:eyes[1,0]+eyes[1,2]]
#le=cv2.cvtColor(le, cv2.COLOR_BGR2GRAY)
re=cv2.cvtColor(re, cv2.COLOR_BGR2GRAY)
re= cv2.resize(re, (21,21),cv2.INTER_AREA)
#lstate= match(refs, le)
rstate= match(refs, re)
if rstate == 'brick':
cv2.putText(img,'closed',(10,90),cv2.FONT_HERSHEY_SIMPLEX,1,(255,0,255),2,cv2.LINE_AA)
## elif lstate == 'brick' or rstate == 'brick':
## cv2.putText(img,'closed/open',(900,900),cv2.FONT_HERSHEY_SIMPLEX,1,(255,0,255),2,cv2.LINE_AA)
else:
cv2.putText(img,'open',(10,90),cv2.FONT_HERSHEY_SIMPLEX,1,(255,0,255),2,cv2.LINE_AA)
cv2.imshow('frame',img)
k= cv2.waitKey(1)
if k == 32:#when u press spacebar
cap.release()
cv2.destroyAllWindows()
Output:
Algorithm used:
Extracting Eye Module
Input: Video /Camera
Method:
For each frame do
1. Detect the faces using Viola & Jones face detection
algorithm
2. Crop the face
3. Detect the eyes in the cropped face using Viola & Jones
eye detection algorithm
for each eye_bounding_box returned
find the pair of eye_bounding_boxes that are
approximately equal in position in y-direction
4. Crop the right eye
Output: eye_image
Eye State Detection Module
Input:
open and closed eye training images and test eye image
Method:
1. Measure a collection of LBPs for each eye image in training and test data
2.Using the histogram (equal-width
bins)of LBP collections of each eye image in training
and test images, calculate the value of the normalized probability density function at the bin.
3. Calculate the distance between each training image's probability
distribution with the probability
distributions of test image's using
Kullback-Leibler Divergence.
4.Output the one with least distance.
Output:
State of the eye whether ‘closed’ or ‘open’
