Robert Cudmore    archive    tags    search    software


Use OpenCV to acquire video

Goals

Install opencv

sudo apt-get update
sudo apt-get upgrade
sudo apt-get install libopencv-dev python-opencv

Check opencv install

pi@pi40 ~ $ /usr/bin/python  
Python 2.7.3 (default, Mar 18 2014, 05:13:23) 
[GCC 4.6.3] on linux2  
Type "help", "copyright", "credits" or "license" for more information.  
>>> import cv2  
>>> cv2.__version__  
'2.4.1'

Install USB thumb drive + usb camera

Got some python/opencv code to acquire

To do

Here is the code

#/usr/bin/python

import numpy as np
import cv2

# initialize camera
cap = cv2.VideoCapture(0)
if cap is None:
    print 'Warning: unable to access camera'

# stopPin = 7 # listen for a change on this pin to stop

# number of frames to acquire
numFrames = 30
currFrame = 0

# outfilebase = 'images/image' # append number + .tif

print 'starting capture of ' + str(numFrames) + ' frames.'
print 'press ctrl-c to stop ... eventually we will stop on a TTL'
while(cap.isOpened()):
    try:
        ret, frame = cap.read()
        if ret==True:
            # flip frame
            # frame = cv2.flip(frame,0)

            # print frame.shape

            # write the flipped frame
            try:
                # out.write(frame)
                # cv2.SaveImage('file' + str(currFrame) + '.tif', frame)
                cv2.imwrite('images/image' + str(currFrame) + '.tif', frame)
            except:
                print 'Error in out.write() OR user hit ctrl-c'
                break

            currFrame += 1
            # if (currFrame > numFrames):
            #    print 'finished ' + str(numFrames) + ' frames.'
            #    break
    
        else:
            break

    except KeyboardInterrupt:
        print 'user hit ctrl-c'
        print 'frames=' + str(currFrame)
        break

# Release camera
cap.release()
Tags: data acquisition

©2017. Robert Cudmore.