Real-world Applications:

• Web graphics, mobile photography
• Medical imaging systems
• Satellite imagery processing
• Social media platforms

Mathematical Foundation

• Image function: $f(x, y)$
• Spatial coordinates: $x \in [0, M-1], y \in [0, N-1]$
• Intensity levels: $L = 2^k$ (k = bit depth)
• Pixel values: $0 \leq f(x,y) \leq L-1$

Digital Characteristics

• Discrete: Finite pixel locations
• Quantized: Limited intensity levels
• Rectangular: Grid-based sampling
• Finite: Bounded image dimensions

Common Bit Depths & Ranges

• 1-bit: Binary (0-1)
  File size = M × N bits
• 8-bit: Grayscale (0-255)
  File size = M × N bytes
• 24-bit: RGB Color (16.7M colors)
  File size = 3 × M × N bytes

1. Binary Images (1-bit)

• Values: 0 (black), 1 (white)
• Applications: OCR, barcode scanning
• Size: Smallest storage requirement

2. Grayscale Images (8-16 bit)

• 8-bit: 256 gray levels (0-255)
• 16-bit: 65,536 levels (medical imaging)
• Size: 1/3 of equivalent color image

3. Color Images (24-32 bit)

• RGB: 8 bits per channel (16.7M colors)
• RGBA: Additional alpha channel
• Size: Largest uncompressed storage

4. Indexed Color (8-bit)

• Palette-based: 256 colors max
• Color lookup table (CLUT)
• Applications: GIF animations, icons

Technical Specifications

• Pixel values: $f(x,y) \in \{0, 1\}$
• Storage: $\frac{M \times N}{8}$ bytes
• Memory efficiency: 8:1 vs grayscale
• Bit packing: 8 pixels per byte

Creation Methods

• Thresholding: $f(x,y) = \begin{cases} 1 & \text{if } g(x,y) > T \\ 0 & \text{otherwise} \end{cases}$
• Dithering: Floyd-Steinberg algorithm
• Edge detection: Canny, Sobel operators

Applications

• OCR systems
• Barcode reading
• Medical bone analysis
• Industrial inspection

Bit Depth Comparison

• 8-bit: 0-255, standard photography
• 12-bit: 0-4095, digital cameras
• 16-bit: 0-65535, scientific imaging
• 32-bit: Floating point, HDR

RGB Conversion

• Luminance: $Y = 0.299R + 0.587G + 0.114B$
• Average: Gray = $\frac{R+G+B}{3}$

Applications

• Medical X-rays
• Satellite imagery
• Security systems
• Scientific imaging

Technical Specifications

• Each channel: 0-255 (8-bit)
• Total colors: $256^3 = 16,777,216$
• Storage: $3 \times M \times N$ bytes
• Memory layout: Interleaved/planar

Extended Formats

• RGB565: Mobile displays (16-bit)
• RGB888: Standard 24-bit
• ARGB8888: 32-bit with alpha
• RGB48: Professional (16-bit/channel)

CMYK (Subtractive Model)

• Channels: Cyan, Magenta, Yellow, Black
• Application: Printing industry
• Color mixing: Ink absorption
• Smaller gamut than RGB

HSV/HSB (Perceptual Model)

• Hue: Color wheel (0-360°)
• Saturation: Purity (0-100%)
• Value: Brightness (0-100%)
• Intuitive color selection

Specialized Spaces

• YUV: Video compression (luminance/chrominance)
• LAB: Perceptually uniform
• XYZ: CIE standard foundation

RGB to HSV Conversion

Technical Structure

• Pixel values: 0-255 (indices)
• Palette: Up to 256 colors
• Storage: $M \times N + 256 \times 3$ bytes
• Effective when colors < 256

Palette Algorithms

• Median Cut: Recursive division
• Octree: Tree-based quantization
• K-means: Color space clustering

Applications

• GIF animations
• Retro games
• Scientific visualization
• Embedded systems

Trade-offs

• + Reduced file size
• + Fast palette swaps
• - Limited color range
• - Posterization effects

Format Classification

• Raster vs Vector: Pixel-based vs mathematical shapes
• Compressed vs Uncompressed: File size optimization
• Lossless vs Lossy: Quality preservation approach
• Single vs Multi-frame: Static images vs animations

Key Selection Criteria

• File Size: Storage and bandwidth requirements
• Quality: Acceptable degradation level
• Compatibility: Software and device support
• Features: Transparency, animation, metadata

Common Format Categories

• Uncompressed: BMP, PPM
  Maximum quality, large files
• Lossless Compressed: PNG, GIF, TIFF
  Quality preserved, moderate compression
• Lossy Compressed: JPEG, WebP
  High compression, quality trade-offs
• Professional: RAW, TIFF, PSD
  Maximum flexibility and quality

Technical Specifications

• File Structure: Header + Pixel data
• Color Support: 1, 4, 8, 16, 24, 32-bit
• Compression: Usually none (RLE available)
• Max Dimensions: Limited by 32-bit integers
• Byte Order: Little-endian (Intel format)

File Structure Details

• File Header: 14 bytes (signature, file size)
• Info Header: 40+ bytes (dimensions, bit depth)
• Color Palette: Optional for ≤8-bit images
• Pixel Data: Row-by-row, bottom-to-top
• Row Padding: Aligned to 4-byte boundaries

Advantages

• Simple structure
• No compression artifacts
• Universal Windows support
• Fast loading

Disadvantages

• Very large file sizes
• Limited web support
• No transparency support
• Minimal metadata

Compression Pipeline

1. Color Space: RGB → YCbCr conversion
2. Subsampling: Reduce chrominance resolution
3. DCT Transform: 8×8 block frequency analysis
4. Quantization: Discard high-frequency data
5. Huffman Coding: Entropy-based compression

Quality Control

• Quality Setting: 1-100 scale (higher = better)
• Quantization Tables: Control frequency removal
• Compression Ratio: 10:1 to 50:1 typical
• Artifacts: Blocking, ringing, mosquito noise

Technical Features

• Progressive: Interlaced loading capability
• EXIF Data: Camera metadata support
• Color Profiles: ICC profile embedding

Advantages

• Excellent compression ratios
• Universal support
• Adjustable quality
• Suitable for photographs

Disadvantages

• Lossy compression
• Compression artifacts
• Poor for text/graphics
• No transparency

Technical Specifications

• Color Types: Grayscale, RGB, Palette, Grayscale+Alpha, RGB+Alpha
• Bit Depths: 1, 2, 4, 8, 16 bits per channel
• Compression: DEFLATE algorithm (LZ77 + Huffman)
• Interlacing: Adam7 progressive display
• Transparency: Full alpha channel support

PNG Chunk Structure

• Critical Chunks: IHDR, PLTE, IDAT, IEND
• Ancillary Chunks: tEXt, gAMA, cHRM, tIME
• CRC Protection: Error detection for each chunk

Compression Techniques

• Filtering: Pre-processing for better compression
• DEFLATE: Dictionary-based compression
• Palette Optimization: Reduce colors when possible

Advantages

• Lossless compression
• Full transparency support
• Robust error detection
• Wide format support

Disadvantages

• Larger than JPEG for photos
• No animation support
• Limited metadata

GIF Characteristics

• Color Limit: 256 colors per frame
• Compression: LZW lossless algorithm
• Animation: Multiple frame support
• Transparency: 1-bit (on/off only)
• Interlacing: Progressive display

TIFF Features

• Flexibility: Multiple compression options
• Color Support: 1-bit to 32-bit per channel
• Metadata: Extensive tag system
• Multi-page: Document scanning support
• Compression: None, LZW, ZIP, JPEG, PackBits

Types of Redundancy

• Spatial: Neighboring pixels often similar
• Spectral: Color channels correlated
• Temporal: Video frame similarities
• Psychovisual: Human eye limitations

Compression Classification

• Lossless: Perfect reconstruction possible
• Lossy: Some information permanently lost
• Near-lossless: Small controlled loss
• Visually lossless: Imperceptible quality loss

Performance Metrics

• Compression Ratio: Original size / Compressed size
• Bit Rate: Bits per pixel (bpp)
• PSNR: Peak Signal-to-Noise Ratio
• SSIM: Structural Similarity Index

Huffman Coding

• Principle: Variable-length codes based on frequency
• Algorithm: Build binary tree from symbol probabilities
• Efficiency: Optimal for symbol-by-symbol coding
• Applications: JPEG (AC coefficients), PNG

LZW (Lempel-Ziv-Welch)

• Principle: Dictionary-based pattern matching
• Algorithm: Build dictionary during encoding
• Advantage: Adaptive, no prior knowledge needed
• Applications: GIF, TIFF, PostScript

Run-Length Encoding (RLE)

• Principle: Replace repeated values with count
• Example: AAABBBBCC → 3A4B2C
• Efficiency: Excellent for binary images
• Applications: BMP, PCX, some TIFF variants

Other Methods

• Arithmetic Coding: Better than Huffman
• DEFLATE: LZ77 + Huffman (PNG)
• LZMA: Advanced dictionary compression

Transform-Based Compression

• DCT (Discrete Cosine Transform): JPEG standard
• Principle: Convert spatial to frequency domain
• 8×8 Blocks: Process image in small segments
• Quantization: Remove high-frequency components
• Entropy Coding: Huffman encode remaining data

Wavelet-Based Compression

• JPEG 2000: Modern wavelet standard
• Multi-resolution: Analyze multiple scales
• Better Quality: Less blocking artifacts
• Progressive: Quality/resolution scalability
• Applications: Medical, satellite imagery

Quantization Strategies

• Uniform: Equal step sizes
• Non-uniform: Perceptually-based steps
• Adaptive: Content-dependent adjustment

Advantages

• High compression ratios
• Adjustable quality
• Suitable for photos

Disadvantages

• Irreversible loss
• Compression artifacts
• Quality degradation

Next-Generation Formats

• WebP: Google's VP8-based format
• HEIF: High Efficiency (H.265-based)
• AVIF: AV1-based open standard
• JPEG XL: Latest JPEG committee standard

Performance Comparison

• Compression Ratio: AVIF > HEIF > WebP > JPEG
• Quality (SSIM): JPEG XL > AVIF > HEIF > WebP
• Encode Speed: JPEG > WebP > HEIF > AVIF
• Decode Speed: JPEG > WebP > HEIF > AVIF
• Browser Support: JPEG > WebP > AVIF > HEIF

Specialized Applications

• Medical Imaging: JPEG-LS, JPEG 2000 (lossless)
• Satellite Data: Custom wavelets, CCSDS standards
• Mobile Photography: HEIF with computational photography

Future Trends

• AI-based compression
• Perceptual quality metrics
• Real-time optimization
• Edge computing integration

Medical Imaging Systems

• DICOM Standard: JPEG-LS, JPEG 2000 lossless
• Requirements: No diagnostic information loss
• Storage: Petabytes of patient data
• Transmission: PACS networks, telemedicine
• Example: 16-bit CT scans, 12-bit X-rays

Social Media Platforms

• Facebook: WebP for web, HEIF mobile uploads
• Instagram: Aggressive JPEG compression with AI enhancement
• TikTok: H.264 video, JPEG thumbnails
• Challenges: Billions of images, mobile bandwidth
• Innovation: Neural super-resolution post-processing

E-commerce & Product Photography

• Amazon: Multi-format delivery (WebP, JPEG, AVIF)
• Requirements: Fast loading, zoom capability
• Progressive Loading: Low-quality → high-quality
• Mobile Optimization: Responsive images, lazy loading

Scientific & Satellite Imagery

• NASA: Custom wavelets, lossless compression
• Google Earth: Multi-resolution pyramids
• Weather Services: Real-time compression for radar

Core Concepts Mastered

• Image Types: Binary, grayscale, color, indexed representations
• Storage Formats: BMP, JPEG, PNG, GIF, TIFF characteristics
• Color Models: RGB, YUV, color space conversions
• Compression: Lossless vs lossy trade-offs and algorithms

Critical Decision Framework

• Quality Requirements: Medical = lossless, Web = acceptable loss
• Storage Constraints: Mobile vs server storage capabilities
• Processing Power: Real-time vs offline compression
• Compatibility: Universal support vs cutting-edge efficiency

Practical Guidelines

• Photography: JPEG for sharing, RAW+lossless for archival
• Graphics/Logos: PNG for transparency, SVG for scalability
• Web Development: Progressive enhancement (AVIF→WebP→JPEG)
• Scientific Data: Always preserve original, compress copies