Key Frame Extraction using Differential Evolution

Python scripts that implement DE for extraction of key frames from a video stream.

NOTE: This repository is intended as a backup for my final year project files.

Requirements:

• Python 2.7
• Additional Python libraries:
  • cv2
  • skimage
  • imageio

This repository contains five python scripts:

1. DE_SSIM.py

   This script uses Average SSIM as the fitness function.
   Input: JPEG images (frames from a video)
   Output: GIF image containing all the extracted key frames.

2. DE_Entropy.py

   This script uses Average Entropy Difference as the fitness function.
   Input: JPEG images (frames from a video)
   Output: GIF image containing all the extracted key frames.

3. DE_Euclidean.py

   This script uses Average Euclidean Distance as the fitness function.
   Input: JPEG images (frames from a video)
   Output: GIF image containing all the extracted key frames.

4. DE_LIVE.py

   This script uses Average Euclidean Distance as the fitness function. It extracts key frames from a live video, it will output a GIF image containing the extracted key frames periodically (depending on how many frames to read and number of key frames required).
   Input: JPEG images (output from capture_xframes)
   Output: GIF image containing all the extracted key frames.

5. capture_xframes.py

   This script captures webcam stream and saves the frames as JPEG files.