MJPEG TOOLS

I use this to get live camera feed. You can manipulate the image on the fly with PIL.

everything is processed in memory so it is suposed to be fast.

Requirements :

• Python >= 2.7
• PIL

Installation :

using pip(recommended) :

pip install mjpeg-tools

manual :

clone this project and run sudo python setup.py install

manual (not recommended):

copy the mjpegtools folder and put it un your project folder

Usage :

to grab a single frame from mjpeg and save it to disk :

from mjpegtools import MjpegParser
image = MjpegParser(url='http://path-to-your-camera-mjpeg').serve()
with open('imagename.jpg', 'wb') as im:
  im.write(image.output.read())

to simply grab infinite frames from mjpeg and save it to disk I would do:

from mjpegtools import MjpegParser
from time import time
While True:
  filename = str(time()).replace('.','') + '.' + 'jpeg'
  image = MjpegParser(url='http://path-to-your-camera-mjpeg').serve()
  with open(filename, 'wb') as im:
    im.write(image.output.read())

Website

Use generator. if you are using Django you can use HttpResponse

flask you can use as_flask_mjpeg:

@app.route('/direct-stream')
def stream_direct():
  cam = MjpegParser(url='http://youripadress/videostream.cgi?user=admin&pwd=password&resolution=8&rate=0')
  cam.quality = 20
  return cam.serve().as_flask_mjpeg()

CHANGELOG

DEC 28, 2014

• Merged Pull requests (@inactivist)
• Updated pip.
• version changed to 0.0.13
• Added License file

DEC 13, 2014:

• Fix README examples.

JUL 27:

• Tests
• Travis

JUL 23:

• Fixed , now working also with dlink dcs-900

JUL 17 :

• Added four methods => as_image , as_mjpeg, as_flask_mjpeg, as_flask_image
• Performance improvement
• Trying to be as pep8 possible.

JUL 10 :

• converted in to a package
• server renamed to mjpegtools
• updated readme => usage , installation
• registered on

MILESTONE :

• implement a server (I have it in my private repo)
• option to convert realtime to other format.