Introduction au Codage d'Images Numériques

Introduction to Digital Image Coding

• Digital image coding is the process of representing an image in a digital format using a specific set of instructions and rules.
• This process compresses the amount of data needed to store or transmit the image without significant loss of information.
• Different methods exist, each with varying levels of compression and data loss.

Key Concepts in Digital Image Coding

• Image Resolution: Refers to the level of detail in an image. Higher resolution images contain more data and are typically larger in file size.
• Pixel: Short for "picture element", it's the smallest unit of an image. Each pixel contains information about color, brightness, or intensity.
• Sampling: The process of converting a continuous image into a discrete set of pixels. Critical for representing a continuous real-world image in a digital format.
• Quantization: The process of assigning discrete values to the sampled pixels, usually measured in bits per pixel (bpp or bits/pixel).
• Compression: Techniques for reducing the size of the digital image file by reducing redundancy and similarity between pixels or regions of pixels. This is a key element of efficient image storage and transmission.

Lossless Compression Techniques

• Run-length encoding (RLE): Repeated sequences of similar pixels are compressed into a code designating the pixel value and the number of times it repeats.
• Huffman coding: Variable-length codes are assigned to pixel values or groups of pixel values based on their frequency.
• Lempel-Ziv (LZW): Code groups of repeated pixel patterns, reducing redundancy based on previously seen sequences.
• Arithmetic coding: Assigns shorter codes to more frequent pixel values and assigns longer codes to less frequent pixel values. More efficient than Huffman coding in many cases.

Lossy Compression Techniques

• JPEG (Joint Photographic Experts Group): Dominant for photographic images. Uses block-based Discrete Cosine Transform (DCT) with quantization to remove less visually important information.
• PNG (Portable Network Graphics): Maintains lossless quality for images with large areas of solid colours, logos, or graphics. Primarily used for non-photographic images in web contexts.
• WebP: Developed by Google, WebP combines lossless and lossy compression, typically compressing images more effectively than JPEG and PNG.
• HEIF (High Efficiency Image Format): Provides higher compression compared to JPEG, often used in mobile devices.

Image Formats and Applications

• Formats like BMP, GIF, TIFF are less commonly used for their inefficient compression compared to newer standards.
• The choice of a digital image coding method depends on factors like the nature of the image, desired file size, and preservation of image quality.
• Medical imaging, scientific imaging, remote sensing, and multimedia applications all make use of various digital image coding techniques.

Discrete Cosine Transform (DCT)

• DCT is a crucial component in lossy image compression techniques like JPEG.
• A mathematical transformation that converts a signal from the spatial domain to the frequency domain.
• DCT analysis of pixel groups allows significant redundancy in image data to be exposed, enabling efficient compression without significant loss of visual detail.
• Quantization levels applied to DCT coefficients control the extent of data loss in the compression process to improve compression ratio.

Quantization

• The process of rounding pixel values to a finite number of levels, usually smaller numbers of levels relative to the original values.
• This data reduction lowers the file size directly by compromising image precision, an inherent characteristic of lossy compression when applied to image data.
• Important to note that quantization levels directly affect the perceived quality of the compressed image.

Ce quiz explore les concepts fondamentaux du codage d'images numériques. Il couvre des thèmes tels que la résolution d'image, les pixels, l'échantillonnage et la quantification. Testez vos connaissances sur la manière dont les images peuvent être compressées et représentées numériquement.