EE200 Signal Processing Project

This repository contains a Jupyter Notebook (ee200_231035(2).ipynb) that demonstrates practical applications of signal processing in two distinct domains: Image Processing and Audio Signal Processing.

The project applies Fourier Transforms, filter design, and spectral analysis to manipulate media files programmatically.

📂 Project Structure

The notebook is divided into two main sections:

1. Hybrid Images (Image Processing)

This section creates a "Hybrid Image"—an optical illusion that looks like one object from up close and another from far away.

• Technique: Uses 2D Fast Fourier Transforms (FFT) to manipulate spatial frequencies.
• Process:
  • Low-pass filter: Applied to the "Dog" image to retain only low frequencies (smooth, blurry shapes).
  • High-pass filter: Applied to the "Cat" image to retain only high frequencies (sharp details/edges).
  • Combination: The frequency domains are merged and converted back to the spatial domain using inverse FFT.
• Result: A hybrid image where the "Dog" appears from a distance, and the "Cat" appears when viewed closely.

2. Instrument Isolation (Audio Processing)

This section processes a mixed audio track to isolate a specific instrument (Saxophone) using digital filtering.

• Technique: Bandpass filtering using scipy.signal.
• Process:
  • Input: Reads a WAV file (song_with_2piccolo.wav).
  • Filter Design: Implements a 4th-order Butterworth bandpass filter with cutoff frequencies at 1150 Hz and 1800 Hz to target the saxophone's frequency range.
  • Spectral Analysis:
    • Plots the Power Spectral Density (PSD) to compare original vs. filtered signals.
    • Generates a Bode Plot to visualize the filter's frequency response.
    • Creates a Pole-Zero Plot for stability analysis.
    • Displays a Spectrogram to visualize frequency changes over time.
• Output: Saves the filtered audio as saxophone_only.wav.

🛠️ Dependencies

To run this notebook, you will need the following Python libraries:

numpy
matplotlib
scipy
Pillow

You can install them using pip:

pip install numpy matplotlib scipy Pillow

🚀 Usage

1. Clone the repository:

   git clone https://github.com/sriramd23/EE200-course-project.git

2. Add Input Files: Ensure the following files are in the same directory as the notebook:
   • cat_gray.jpg (High-frequency source)
   • dog_gray.jpg (Low-frequency source)
   • song_with_2piccolo.wav (Audio source)
3. Run the Notebook: Open ee200_231035(2).ipynb in Jupyter Notebook or Google Colab and run all cells.

📊 Outputs

The notebook generates the following outputs:

• Visualizations:
  • Hybrid Image and Frequency Spectra.
  • PSD Comparison Plot (psd_comparison.png).
  • Bode Plot (bode_plot.png).
  • Pole-Zero Plot (pole_zero_plot.png).
  • Spectrogram (spectrogram.png).
• Audio:
  • saxophone_only.wav: The processed audio file containing the isolated instrument.

👤 Author

sriram dhanasekaran Course: EE200