Timelapse Maker

Was trying to find a Mac timelapse maker that creates timelapses from screenshots and not from recording a video of your screen. This is my simple, naive attempt at it.

Timelapse-maker captures screenshots at regular intervals and then stitches them together to create a timelapse video. It uses the pyautogui library for capturing screenshots and the cv2 library for video processing.

Not tested on Windows, but open to PRs and user feedback!

Requirements

• Python 3.x
• pyautogui library
• cv2 library

Usage

1. Clone the repository and navigate to the project directory.

2. Install the required dependencies using the following command:

   pip install pyautogui opencv-python
   

3. The script accepts the following command-line arguments:

   • --num_hours: Number of hours to capture screenshots (default is 2).
   • --fps: Frames per second for the timelapse video (default is 15).

4. Run the script using the following command:

   python timelapse_maker.py --num_hours <hours> --fps <fps>
   

   Replace <hours> with the desired number of hours for capturing screenshots, and <fps> with the desired frame rate for the timelapse video.

5. The script will capture screenshots for the specified duration and save them in a directory with a timestamp.

6. After capturing the screenshots, the script will create a timelapse video by stitching the images together.

7. Once the video is created, it will be saved in the project directory with a timestamp.

Notes

• The screenshots will be captured at one-second intervals by default. You can modify the interval by adjusting the time.sleep() function in the capture_screenshots() function.

• The timelapse video will be saved in the MP4 format with the specified frames per second (FPS).

• Make sure to have a sufficient amount of disk space as capturing screenshots and creating the video may require significant storage depending on the duration and frequency of screenshots.

• The script uses the current timestamp to generate unique directory and file names for each run, ensuring that previous results are not overwritten.

• Feel free to modify the code to suit your specific needs.

License

This project is licensed under the MIT License.