DCT & Frequency Domain Steganography

Spatial vs. Frequency Domain Steganography

• Spatial Domain (PNG / BMP / WAV): Data is hidden directly in pixel RGB color values or raw PCM audio samples (e.g., LSB replacement). Easily detected by histogram analysis (zsteg).
• Frequency Domain (JPEG / MP3): Data is converted into frequency coefficients using mathematical transforms (Discrete Cosine Transform (DCT) or Fast Fourier Transform (FFT)) before lossy compression.

---

1. JPEG DCT Steganography (Jsteg)

JPEG compression divides images into $8 \times 8$ pixel blocks, applies DCT to convert pixels into frequency coefficients, and quantizes the coefficients.

Jsteg embeds secret data into the LSBs of non-zero quantized DCT coefficients:

# Extract hidden data using jsteg
jsteg reveal input.jpg extracted_secret.txt

# Embed payload into JPEG
jsteg hide input.jpg payload.txt output_stego.jpg

---

2. OutGuess Steganography

OutGuess is a universal steganographic tool that embeds secret data into redundant bits of coefficients while preserving original statistics and coefficient histograms to defeat steganalysis detection algorithms.

# Extract hidden message with optional password
outguess -k "secretpass" -r input.jpg extracted_output.txt

---

3. StegDetect & StegBreak

stegdetect automates statistical detection of steganographic content in JPEG files (Jsteg, OutGuess, JPHide, F5):

# Scan JPEG images for stego signatures
stegdetect -s 10.0 *.jpg

# Brute-force dictionary attack against OutGuess or JPHide
stegbreak -r rules.ini -f wordlist.txt -t jphide input.jpg