Lapse reduces the hassle of setting up a time-lapse shoot from a headless Raspberry Pi.
Lapse is a command-line app that provides UI that can be accessed from a browser elsewhere on your intranet.
Lapse is a working rough draft. I'm developing it to scratch a personal itch. At the moment, stitching pics together to make the time-lapse video is done manually. Eventually, Lapse will invoke mencoder.
The presentation needs work.
Pull Requests for bugs will be appreciated; PRs for features might get ignored. You're welcome to fork this and head off in your own direction.
The latest Raspian includes the picamara packages.
pip install flask
gets you everything else you need.
python lapse.py
starts the application. Issuing /quit through a browser stops it.
Lapse can run in a Python virtual environment. It requires OpenCV and the cv2 Python bindings. To get those for Python 2.7:
sudo apt-get install python-dev python-opencv libjpeg-dev
(At present, using Python 3.x requires building OpenCV manually.)
To make cv2 available inside the virtual environment, create the environment via
virtualenv --system-site-packages venv
. venv/bin/activate
pip install -r requirements.txt
Then,
python lapse.py
starts the app. From a browser elsewhere on your intranet, visit http://yourpi:5000/
Clicking on the image will refresh it, allowing you to adjust the camera position to your liking before starting collection.
Path Prefix is the path to where images will be saved, plus a prefix. The path must exist. A timestamp and ".jpg" are appended to the prefix for each image saved.
Delay is the amount of time to delay before starting to record. A simple number is interepreted as seconds. Appending 'h', 'm', or 's' specifies hours, minutes, and seconds, respectively. These can combined. 1h 30m will delay for an hour and a half before beginning to record.
Record is the length of time to record. 12h will record for half a day.
Sec/Pic is the the number of seconds between each picture. A Raspberry Pi 3 is just fast enough to record 1 medium-sized JPEG per second to fast media (e.g., a good class 10 SD card or a fast thumb drive). If you notice stuttering in the final video, try a larger number.
Make sure you have sufficient storage for the number of images that you'll be capturing.
Click Start to begin.
After the Delay + Record time has elapsed, quit the lapse via /quit from a remote browser, and use mencoder or the tool of your choice to convert the images into a video.
There are some settings in lapse.py that may need manual adjustment. Consult the code. These may be exposed through the UI in the future.
I borrowed from several sources. Credits are in the source.
MIT