Texture Mapping

UV mapping is the process of projecting a 2D texture image onto a 3D model's surface. It involves creating a UV unwrap, where the 3D mesh is 'flattened' into 2D space. Each vertex on the 3D model is assigned a corresponding 2D coordinate (U, V) on the texture image. This creates a mapping that tells the rendering engine which texel (texture pixel) to apply to which part of the polygon.

• UV coordinates range from 0 to 1 across the texture space.
• A poorly unwrapped UV map can cause visible stretching or seams in the final render.
• This is a foundational, manual, or semi-automated step in the 3D asset creation pipeline.

Beyond simple color (albedo/diffuse map), specialized texture maps encode different surface properties for physically based rendering (PBR):

• Normal Maps: Encode surface normal perturbations in RGB channels to simulate fine geometric detail (e.g., bumps, scratches) without adding polygons.
• Specular/Roughness Maps: Control the shininess and micro-surface detail of a material. A roughness value of 0 is perfectly mirror-like, while 1 is completely matte.
• Displacement Maps: Actually modify the geometry of the mesh by displacing vertices along their normals, requiring a highly tessellated base mesh.
• Ambient Occlusion (AO) Maps: Pre-calculate and bake soft shadowing into crevices and contact points to enhance depth perception.
• Emissive Maps: Define which parts of a surface emit their own light.

When a textured surface is viewed from a distance, a single screen pixel may cover many texture pixels (texels), causing aliasing and flickering (moire patterns). Mipmapping solves this by pre-calculating a chain of progressively lower-resolution versions of the original texture.

The rendering engine automatically selects the appropriate mip level based on the surface's distance and size on screen. Texture filtering (bilinear, trilinear, anisotropic) is then applied to the selected mip level to smooth the final sampled color. Anisotropic filtering is a more advanced technique that accounts for surfaces viewed at oblique angles, preserving detail better than standard trilinear filtering.

Instead of using a bitmap image, procedural texturing generates texture patterns algorithmically using mathematical functions and noise (e.g., Perlin, Worley noise). Key advantages include:

• Infinite Resolution: Textures are calculated on-the-fly and have no fixed resolution, avoiding pixelation at any zoom level.
• Parametric Control: Artists can adjust material properties (e.g., scale, contrast, color) via sliders rather than repainting bitmaps.
• Memory Efficiency: A small algorithm can describe a vast, complex texture, reducing GPU memory footprint.
• Seamless Tiling: Patterns can be designed to tile perfectly across any surface. Common uses include terrain generation, organic materials like wood and marble, and sci-fi surfaces.

Texture atlasing is an optimization technique where multiple smaller textures (for different parts of a model or for different models) are packed into a single, larger texture sheet.

• Reduces Draw Calls: The GPU can render many objects with different textures in a single batch by binding just one atlas texture, significantly improving rendering performance.
• Improves Cache Efficiency: Accessing texels from a contiguous memory region is faster for the GPU.
• Packing Algorithms: The process of arranging the sub-textures into the atlas to maximize space utilization is a complex bin-packing problem, often solved with tools like MaxRects or Guillotine algorithms. This is essential for real-time applications like games and mobile AR.

Beyond standard UV mapping, several techniques address specific visual challenges:

• Triplanar Projection: Projects a texture along the three world-space axes (X, Y, Z) and blends them based on surface normals. This eliminates the need for UV unwrapping and prevents seams on complex organic shapes, commonly used for terrain.
• Parallax Occlusion Mapping: A screen-space technique that uses a height map to offset texture coordinates, creating a convincing illusion of depth and parallax on flat surfaces.
• Texture Bombing: Instances small texture elements (e.g., rocks, grass clumps) randomly across a surface according to a density map, creating varied, non-repeating detail.
• Volume Texturing: Applies 3D textures (noise, gradients) within a volume, used for rendering materials like smoke, marble, or wood where the internal structure is visible.

The Diffuse Map (or Albedo Map) defines the base color of a surface, representing how it reflects light uniformly across all viewing angles, ignoring specular highlights. It contains the object's intrinsic color without lighting or shadow information.

• Purpose: Provides the fundamental color data for a material.
• Key Property: Should be lit-neutral; shadows and baked lighting are artifacts to be avoided.
• Example: The red color of a brick wall or the green of a leaf.

A Normal Map is a texture that encodes surface normal direction perturbations in RGB channels, simulating high-frequency geometric detail on a low-polygon mesh without altering the underlying 3D geometry.

• Mechanism: Each pixel's (R, G, B) value corresponds to a (X, Y, Z) vector in tangent space.
• Visual Effect: Creates the illusion of bumps, dents, and fine surface wrinkles by affecting per-pixel lighting calculations.
• Limitation: Does not alter the object's silhouette or cast shadows onto other objects.

These maps control a surface's interaction with direct light sources. A Specular Map defines the intensity and color of specular highlights (common for metals). A Roughness Map (or Glossiness Map) defines microsurface detail, controlling how sharp or blurred reflections appear.

• Roughness Map: Black (0.0) represents a perfectly smooth, mirror-like surface. White (1.0) represents a completely rough, matte surface.
• Workflow: Modern Physically Based Rendering (PBR) pipelines typically use a metallic/roughness workflow, combining a Metallic map with a Roughness map.

Displacement Maps and Height Maps are grayscale textures where pixel intensity represents height. Unlike normal maps, they actually modify the geometry of the base mesh at render time.

• Displacement Mapping: Tessellates the base mesh and physically displaces vertices along their normals. This changes the silhouette and can cast self-shadows.
• Height Map: A simpler form often used in parallax occlusion mapping, a screen-space technique that creates a convincing illusion of depth.
• Compute Cost: Significantly more expensive than normal mapping due to geometry alteration.

An Ambient Occlusion (AO) Map is a grayscale texture that simulates soft shadowing in crevices and areas where ambient light is occluded. It adds contact shadows and depth, enhancing the perception of geometric complexity.

• Purpose: Pre-computes and bakes global illumination-like soft shadows that are independent of direct lighting.
• Usage: Multiplied over the diffuse/albedo color in the shader.
• Baked vs. Real-Time: Often baked from a high-poly model but can also be approximated in real-time via Screen-Space Ambient Occlusion (SSAO) techniques.

An Emission Map defines areas on a surface that emit their own light, independent of scene lighting. It is typically an RGB texture where black (0,0,0) indicates no emission and brighter values define the color and intensity of the emitted light.

• Application: Used for materials like LED screens, neon signs, bioluminescent organisms, or control panels.
• Rendering: Adds light contribution to the scene in global illumination or real-time lighting engines, often via a post-process bloom effect for glow.
• Key Property: Does not require external light sources to be visible.

The process of creating a polygonal surface representation—a mesh—composed of vertices, edges, and faces. This is the foundational 3D geometry to which textures are applied. Common outputs include:

• Triangle meshes: The standard for real-time rendering.
• Quad meshes: Preferred for subdivision surfaces and animation.

It converts raw 3D data from sources like point clouds or TSDF volumes into a renderable format.

The science of obtaining reliable measurements and 3D reconstructions from photographs. It encompasses the full pipeline from images to textured 3D models.

Key techniques include:

• Structure from Motion (SfM): Recovers sparse 3D points and camera poses.
• Multi-View Stereo (MVS): Generates dense geometry (point clouds/meshes).

The final output is often a mesh with photorealistic textures automatically projected from the input images.

A specialized texture that stores surface normal direction vectors in the RGB color channels. It simulates high-resolution surface detail—like bumps, grooves, and scratches—on a low-polygon model without altering the underlying geometry.

• Mechanism: Perturbs the surface normals used in lighting calculations.
• Use Case: Essential for adding fine detail while maintaining performance in games and real-time applications.
• Creation: Often baked from a high-poly model or generated from height maps.

The process of inferring underlying scene properties from 2D observations. It reverses the traditional graphics pipeline to estimate:

• Geometry (the 3D shape)
• Material (e.g., albedo, roughness, metallic)
• Lighting conditions

This is a core research area for creating digital twins and enabling relighting of captured objects. Differentiable rendering is a key enabling technology, allowing gradient-based optimization of these parameters.

A 3D scanning method that uses synchronized color (RGB) and depth (D) images from sensors like Intel RealSense or Microsoft Kinect. The depth data provides direct 3D measurements, enabling real-time reconstruction.

Common pipeline:

1. Depth frames are fused into a volumetric representation like a Truncated Signed Distance Function (TSDF).
2. A mesh is extracted via Marching Cubes.
3. RGB frames are used to create a color texture for the mesh.

This is fundamental for real-time 3D mapping in robotics and AR.

The process of flattening a 3D mesh's surface into a 2D plane to create a UV map. This 2D representation defines how the pixels of a texture image correspond to specific points on the 3D geometry.

• Challenge: Minimizing distortion and seams while efficiently using texture space.
• Result: A set of 2D coordinates (U,V) for each mesh vertex.

This is a critical, often manual, step performed in 3D modeling software before texture painting or application.