I am trying to run some python 2.7 code with opencv2. Currently if a subject enters the frame the code labels them as the one it looks like most in its photo database. Instead of this I would like it to label untrained faces as "unknown" this is my code so far:
import cv2
import numpy as np
faceDetect=cv2.CascadeClassifier('haarcascade_frontalface_default.xml');
cam=cv2.VideoCapture(0);
rec=cv2.createLBPHFaceRecognizer();
rec.load("recognizer\\trainingData.yml")
id=0
font=cv2.cv.InitFont(cv2.cv.CV_FONT_HERSHEY_SIMPLEX,1,1,0,0)
#font=cv2.cv.InitFont(cv2.cv.CV_FONT_HERSHEY_COMPLEX_SMALL,3,1,0,1)
while (True):
ret, img=cam.read();
gray=cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
faces=faceDetect.detectMultiScale(gray,1.3,5);
for (x,y,w,h) in faces:
#cv2.rectangle(img,(x,y),(x+w,y+h),(0,255,255),2)
id,conf=rec.predict(gray[y:y+h, x:x+w])
if(id==1):
id="Admin"
cv2.rectangle(img,(x,y),(x+w,y+h),(0,255,255),2)
elif(id==2):
id="Sonja"
cv2.rectangle(img,(x,y),(x+w,y+h),(255,255,255),2)
cv2.rectangle(img,(x,y),(x+w,y+h),(255,255,255),2)
cv2.cv.PutText(cv2.cv.fromarray(img),str(id),(x,y+h),font,(255,255,255));
cv2.imshow("Image",img);
if(cv2.waitKey(1)==ord('q')):
break;
cam.release()
cv2.destroyAllWindows()
And the trainer is run through this code:
import os
import cv2
import numpy as np
from PIL import Image
recognizer=cv2.createLBPHFaceRecognizer();
path='dataSet'
def getImagesWithID(path):
imagePaths=[os.path.join(path,f) for f in os.listdir(path)]
faces=[]
IDs=[]
for imagePath in imagePaths:
faceImg=Image.open(imagePath).convert('L');
faceNp=np.array(faceImg,'uint8')
ID=int(os.path.split(imagePath)[-1].split('.')[1])
faces.append(faceNp)
print ID
IDs.append(ID)
cv2.imshow("training",faceNp)
cv2.waitKey(10)
return np.array(IDs), faces
Ids, faces= getImagesWithID(path)
recognizer.train(faces, Ids)
recognizer.save('recognizer/trainingData.yml')
cv2.destroyAllWindows()
I assume that your rec.predict method returns the id of the recognized faces. In which case you can add an 'else' condition in the for (x,y,w,h) in faces: loop for the faces that don't have an id, like so:
if(id==1):
id="Admin"
cv2.rectangle(img,(x,y),(x+w,y+h),(0,255,255),2)
cv2.putText(img,"Admin",x+h/2,y+w+40),cv2.FONT_HERSHEY_SIMPLEX,0.6,(0,255,255),2)
elif(id==2):
id="Sonja"
#cv2.rectangle(img,(x,y),(x+w,y+h),(255,255,255),2)
cv2.rectangle(img,(x,y),(x+w,y+h),(255,255,255),2)
cv2.putText(img,"Sonja",x+h/2,y+w+40),cv2.FONT_HERSHEY_SIMPLEX,0.6,(255,255,255),2)
else:
cv2.rectangle(img,(x,y),(x+w,y+h),(0,0,255),2)#Red
cv2.putText(img,"Unknown",(x+h/2,y+w+40),cv2.FONT_HERSHEY_SIMPLEX,0.6,(0,0,255),2)
Also, I've implemented the putText method differently than the one you've shown.
If you print the value of conf you can see that the value will be less than 70 if the face is in the dataset you made otherwise the value of conf will be more than 70. The value 70 is set by me if you want the system be more accurate check the value of conf and give any other value as your wish.
import cv2
import numpy as np
from time import sleep
faceDetect=cv2.CascadeClassifier('haarcascade_frontalface_default.xml')
cam=cv2.VideoCapture(0)
rec=cv2.createLBPHFaceRecognizer()
rec.load("recognizer/trainingData.yml")
id=0
font=cv2.cv.InitFont(cv2.cv.CV_FONT_HERSHEY_COMPLEX_SMALL,1,1,0,1)
while True:
ret,img=cam.read();
gray=cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
faces=faceDetect.detectMultiScale(gray, 1.3, 5)
for(x,y,w,h) in faces:
cv2.rectangle(img, (x,y), (x+w,y+h), (255, 0, 0), 2)
id,conf=rec.predict(gray[y:y+h, x:x+w])
if(conf<70):
if(id==1):
id="Admin"
if(id==2):
id="Sonja"
else:
id="unknown"
cv2.cv.PutText(cv2.cv.fromarray(img), str(id), (x,y+h), font,255)
print ('ok')
cv2.imshow("Face",img)
if (cv2.waitKey(1) & 0xFF==ord('q')):
break
cam.release()
cv2.destroyAllWindows()
if conf<70:
if id !=0 & id !=1:
cv2.putText(img,"unknown",(x,y),cv2.FONT_HERSHEY_SIMPLEX,1,(0,255,0),2,cv2.LINE_AA)
I am pulling PNG images from Jupyter Notebooks and manage to display with IPython.display.Image but not with matplotib.pyplot.plt. What am I missing? I use python 2.7.
I am using the following algorithm:
To open the notebook JSON content I do:
import nbformat
notebook_ = nbformat.read(file_notebook, 4)
After retrieving the relevant cell information I pull the png information from it using:
def cell_to_image(cell, out_value_item_number=1):
if "execution_count" in cell.keys(): # i.e version >=4
return cell["outputs"][out_value_item_number]['data']['image/png']
elif "prompt_number" in cell.keys(): # i.e version < 4
return cell["outputs"][out_value_item_number]['png']
return None
cell_image = cell_to_image(cell)
The first few characters of cell_image (which is unicode) looks like:
iVBORw0KGgoAAAANSUhEUgAAA64AAAFMCAYAAADLFeHSAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\n
AAALEgAACxIB0t1+/AAAIABJREFUeJzs3Xd8jef/x/HXyTjZiYQkCGrU3ruR0tr9oq2qGtGo0dbe
\nm5pVlJpFUSMoVb6UoEZ/lCpatWuPUiNEEiMDmef3R75OexonJKUO3s/HI4/mXPd1X/d1f+LRR965
\n7/u6DSaTyYSIiIiIiIiIjbJ70hMQERERERERyYiCq4iIiIiIiNg0BVcRERERERGxaQquIiIiIiIi
\nYtMUXEVERERERMSmKbiKiIiIiIiITVNwFRGRxyIkJIRixYqxfv36+24/e/YsxYoVo3jx4v/yzGxb
\naGgoderUIS4uDoBdu3bRsmVLKlasyCuvvMKgQYOIjo622CcsLIyGDRtSunRp6tSpw8KFC62OW7p0
\naRo2bJju53Lnzh1GjRrFyy+/TNmyZWnRogW//fbbQ835q6++olGjRpQvX5769eszc+ZMkpOTzdtT
\nU1OZNGkSNWrUoHTp0jRp0oTdu3enGyc2NpZOn
I can easily plot in my Jupityer notebook using
from IPython.display import Image
Image(cell_image)
And now to my question:
How can I manipulate cell_image to be plt.subplot friendly?
(Assuming import matplotlib.pyplot as plt).
I realise that plt.imshow wouldn't work because this would require an array, which is not my case (which is a string, as far as I understand).
If you have your image string representation in a variable string_rep, the following code should work.
from io import BytesIO
import matplotlib.image as mpimage
import matplotlib.pyplot as plt
with BytesIO(string_rep.decode('base64')) as byte_rep:
image = mpimage.imread(byte_rep)
plt.imshow(image)
import cv2
import numpy as np
cap = cv2.VideoCapture('traffic.avi')
retval, frame = cap.read()
print retval
================ RESTART: J:\Python For DIP\traffic_video.py ================
False
>>>
The Value of retval is always False, which means the video is not read by the command. It must be True to read frames. I don't know what to do. However when I use my default webcam it turns to be True. I tried many videos and the same problem appears. Note: I have installed the ffmpeg correctly.
Note: This is not the full code, in this step I am only validating cap.read() either True or False
This method is guaranteed 100%
first of all check your version of OpenCV, say for instance 2.4.11. you can check it by typing the following commands in your Python Shell:
>>> from cv2 import __version__
>>> __version__
'2.4.11'
>>>
Then go to C:\opencv\build\x86\vc12\bin and copy opencv_ffmpeg2411.dll.
Finally go to root directory of Python ex: C:\Python27 and paste opencv_ffmpeg2411.dll in it
check the name of the file opencv_ffmpeg2411.dll, whether the version
of opencv is written or not, if not do the following
opencv_ffmpeg(version of your opencv without dots).dll
After that create a new Python file and copy this code and paste it loading your own video
import numpy as np
import cv2
# Capture video from file
cap = cv2.VideoCapture('your video')
while True:
ret, frame = cap.read()
if ret == True:
gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
cv2.imshow('frame',gray)
if cv2.waitKey(30) & 0xFF == ord('q'):
break
else:
break
cap.release()
cv2.destroyAllWindows()
you will have an output video for example like this:
Result
Finding the root directory of Python can be a little tricky. I am using an Enthought distribution and, at first, pasted the opencv_ffmpeg file into the wrong Python directory.
WRONG:
C:\Users\USERNAME\AppData\Local\Programs\Python\Python35-32
RIGHT:
C:\Users\USERNAME\AppData\Local\Enthought\Canopy\User
Long story short, make sure you find the right Python directory.