Project Overview

This project encompasses a range of functionalities related to image and video processing, particularly focusing on the detection and classification of certain animals in images.

Content

1. Image Classification (ia_v1.py & IA.ipynb)

• Utilizes the VGG16 model from TensorFlow's Keras library for image classification.
• Specifically, detects animals such as 'coyote', 'timber_wolf', 'white_wolf', 'red_wolf', and 'Irish_wolfhound'.

2. Video Processing (video.ipynb)

• Splits videos into frames at specified intervals.
• Useful for extracting images from video footage for further analysis.

Dependencies

To set up the project, you'll need the following dependencies:

pip install tensorflow keras opencv-python Pillow

Usage

Image Classification

Use the classify_image(image_path) function, providing the path to your image. The function will return whether one of the specified animals is detected and its associated probability.

Video Processing

To process videos and extract frames at specified intervals, follow these steps:

1. Use the decouper_video function.
2. Provide the video_path which is the path to your video.
3. Specify the output_folder which is the directory where you want the extracted frames to be saved.
4. Set the interval which is the time interval (in seconds) between frames that should be saved.

Example:

video_path = "path_to_your_video.mp4"
output_folder = "desired_output_directory"
interval = 1  # Extracts a frame every 1 second
decouper_video(video_path, output_folder, interval)