Fact-checked by Grok 2 weeks ago

3D projection

3D projection is a fundamental technique in computer graphics and descriptive geometry that transforms three-dimensional coordinates into a two-dimensional representation, allowing the depiction of 3D scenes on flat surfaces like screens or paper.^[1]^[2] The two primary categories of 3D projection are perspective projection and parallel projection.[https://www.geeksforgeeks.org/computer-graphics/projections-in-computer-graphics/)^[3] Perspective projection simulates the way the human eye perceives depth, where objects farther from the viewer appear smaller, and parallel lines converge at vanishing points, providing a realistic visual effect essential for immersive applications like video games and virtual reality.^[1]^[4] A common form of parallel projection is the orthographic projection, which projects points orthogonally onto the viewing plane without perspective distortion, preserving the actual sizes and angles of objects, which makes it ideal for technical drawings, engineering blueprints, and isometric views in architecture.^[1]^[2] Mathematically, 3D projections are implemented using 4x4 transformation matrices in homogeneous coordinates, which map world coordinates through a series of steps in the graphics pipeline: modeling, viewing, projection, and clipping.^[5]^[3] For perspective projection, the matrix incorporates the field of view and aspect ratio to create depth scaling, often represented as:

\begin{pmatrix} \frac{2n}{r-l} & 0 & \frac{r+l}{r-l} & 0 \\ 0 & \frac{2n}{t-b} & \frac{t+b}{t-b} & 0 \\ 0 & 0 & -\frac{f+n}{f-n} & -\frac{2fn}{f-n} \\ 0 & 0 & -1 & 0 \end{pmatrix}

where n, f, l, r, t, and b denote near plane, far plane, left, right, top, and bottom clipping boundaries, respectively.^[5] Orthographic projections use simpler matrices that uniformly scale and translate without depth division.^[4] These projections play a crucial role in the rendering pipeline by converting eye-space coordinates to clip-space, enabling efficient rasterization and display of complex 3D models in fields ranging from animation and simulation to scientific visualization.^[1]^[6]

Fundamentals

Definition and Principles

3D projection is the process of mapping points from a three-dimensional space onto a two-dimensional surface, typically representing a 3D scene or object on media such as paper, screens, or displays by extending lines of sight from a defined viewpoint through the 3D points to intersect a projection plane.^[7] This technique simulates how human vision or a camera captures depth and spatial relationships, converting volumetric data into a flat image while preserving essential geometric properties.^[8] The core principles revolve around the viewpoint, often modeled as an eye or camera position, and the projection plane, which acts as the 2D canvas where the image forms. Rays emanate from the viewpoint through each point in the 3D object; in parallel projections, these rays remain non-converging for uniform scaling, whereas in converging projections, they meet at vanishing points to mimic realistic depth cues like foreshortening.^[7] A 3D object, composed of vertices (points), edges (lines connecting vertices), and faces (polygonal surfaces bounded by edges), maps to corresponding 2D points, lines, and polygons on the plane through this ray intersection process, akin to a simple ray-tracing where each ray traces a straight path from viewpoint to projection surface.^[8] This method plays a vital role in visualization across disciplines: in engineering, it facilitates precise communication of object dimensions and assemblies via technical drawings; in art, it enables the illusion of depth through techniques like linear perspective; and in computing, it underpins rendering in graphics software for simulations and virtual environments.^[9]^[10] By conveying spatial relationships, relative depths, and proportional scaling on a 2D medium, 3D projection bridges the gap between physical or virtual three-dimensionality and accessible two-dimensional representation, with parallel and perspective serving as the two broad categories.^[7]

Historical Development

The origins of 3D projection techniques can be traced to ancient Greece, where early attempts at representing depth and spatial relationships in art emerged around the fifth century BCE, laying groundwork for geometric approaches to visualization.^[11] Euclid of Alexandria advanced these ideas in his treatise Optics, composed around 300 BCE, which provided the first systematic Greek exploration of the geometry of vision and perspective, treating sight as rays emanating from the eye to form visual cones.^[12] During the Renaissance, Filippo Brunelleschi pioneered linear perspective in 1415 through experiments in Florence, using mirrors and painted panels of the Baptistery to demonstrate how converging lines could realistically depict three-dimensional space on a flat surface, rediscovering principles lost since antiquity.^[13] Leonardo da Vinci further integrated mathematics and art in the late 15th century, employing geometric perspective in sketches and paintings to achieve accurate spatial depth, as seen in works like The Last Supper where architectural elements converge to a vanishing point.^[14] Albrecht Dürer popularized these methods in northern Europe with his 1525 treatise Underweysung der Messung, which detailed practical tools and constructions for linear perspective, including devices to aid artists in drawing accurate projections.^[15] In the late 18th century, Gaspard Monge formalized descriptive geometry in the 1790s, developing orthographic projection as a method to represent three-dimensional objects through multiple two-dimensional views, essential for engineering and technical drawing.^[16] This approach shifted focus from artistic illusion to precise mechanical representation, influencing industrial design throughout the 19th century. The 20th century brought computational advancements, beginning with Ivan Sutherland's 1963 Sketchpad system, the first interactive computer graphics program that enabled real-time manipulation of 2D drawings using a light pen, laying the groundwork for digital 3D projection in subsequent developments like Sketchpad III.^[17] In the 1980s, ray tracing emerged in computer-generated imagery (CGI), with Turner Whitted's 1980 algorithm simulating light rays for realistic reflections and shadows, revolutionizing film visuals in productions like those from Pixar.^[18] John Carmack advanced real-time 3D projection in game engines during the 1990s, optimizing techniques like binary space partitioning in Doom (1993) and Quake (1996) to render complex scenes efficiently on consumer hardware.^[19] Post-2010, integrations of 3D projection in virtual reality (VR) and augmented reality (AR) have proliferated, with head-mounted displays like Oculus Rift (2012) and Microsoft HoloLens (2015) employing stereoscopic projections to overlay virtual 3D elements onto real environments, driving immersive applications in training and simulation.^[20] In the 2020s, advancements continued with the release of Apple's Vision Pro in February 2024, a spatial computing device that employs high-resolution stereoscopic 3D projections to blend digital and physical environments, further popularizing immersive AR/VR applications.^[21]

Projection Techniques

Parallel Projection

Parallel projection is a fundamental technique in three-dimensional (3D) visualization and graphics, characterized by projection lines that are all parallel to one another and directed toward the projection plane at a fixed angle, often perpendicular or oblique. This approach positions the viewpoint conceptually at infinity, ensuring that the apparent size of objects remains constant regardless of their distance from the projection plane, thereby achieving uniform scale across the entire image. Unlike methods that simulate human vision, parallel projection produces no vanishing points, as parallel lines in 3D space retain their parallelism in the 2D projection.^[22] The core mechanism relies on rays emanating in a consistent direction from each point on a 3D object, intersecting the projection plane without convergence, which preserves certain distances and angles along the projection direction. This uniform treatment of depth eliminates foreshortening effects, allowing for consistent representation of shapes and dimensions. In the general process, 3D coordinates (x, y, z) are transformed by projecting along a specified direction—typically scaling or ignoring the z-component uniformly—to map onto 2D coordinates, facilitating the creation of flat representations suitable for technical documentation.^[22] Key advantages of parallel projection include its capacity for accurate measurement of dimensions directly from the projected view, with no distortion introduced by depth variations, making it essential for engineering and design tasks requiring precision. It excels in maintaining parallelism, which supports clear multiview representations without the optical illusions of convergence. Subtypes such as orthographic, oblique, and axonometric projections fall under this category, each adapting the fixed-angle principle for specific representational needs.^[23]^[24] In contrast to perspective projection, which introduces realism through converging rays and depth-based scaling to replicate natural viewing, parallel projection prioritizes technical fidelity over visual depth cues, resulting in a more abstract but measurably reliable output ideal for blueprints and schematic drawings in computer-aided design (CAD) systems.^[22]^[25]

Perspective Projection

Perspective projection is a method of representing three-dimensional objects on a two-dimensional plane by simulating the way the human eye or a camera perceives depth, where lines of sight converge from a fixed viewpoint toward the scene. This technique creates a realistic illusion of depth by making objects farther from the viewer appear smaller and parallel lines in three-dimensional space appear to converge at vanishing points on the projection plane. The core mechanism involves rays emanating from a center of projection—analogous to the eye or camera lens—passing through points on the 3D object and intersecting the image plane, resulting in a central perspective with a single vanishing point for lines parallel to the line of sight, or multiple vanishing points in multi-point perspectives for more complex scenes.^[7]^[1] In the general process, the projection plane acts as the picture surface, and the horizon line represents the eye level, where vanishing points lie to depict the convergence of parallel elements like roads or building edges in the third dimension. This introduces foreshortening, where the projected size of an object diminishes with increasing distance from the viewer, enhancing spatial realism. Perspective projection has been foundational in art since the early 15th century, pioneered by Filippo Brunelleschi through experiments with peephole views of Florentine architecture, and formalized in Leon Battista Alberti's treatise Della pittura (1435), which described systematic rules for depth representation in painting. It remains essential in architecture for accurate renderings and in photography, where camera lenses inherently produce perspective views to capture natural visual scenes.^[26]^[10]^[27] The advantages of perspective projection include its ability to convey natural depth perception and relative sizes, making it superior for immersive visualizations compared to parallel projections. It effectively handles the scaling of distant objects, providing a more intuitive sense of scale and distance in representations. A variant known as weak perspective projection approximates full perspective for scenes where objects are relatively far from the viewer and depth variations are small compared to the average distance, treating the projection as a scaled orthographic one; this simplification is commonly used in computer vision tasks to reduce computational complexity while retaining approximate depth cues.^[1]^[28]^[29]

Specific Projection Methods

Orthographic Projection

Orthographic projection is a specific type of parallel projection in which the projection lines, or rays of sight, are perpendicular to the projection plane, resulting in a two-dimensional representation that preserves the true size and shape of the object's features parallel to that plane.^[9] This method positions the projection plane between the observer and the object, ensuring that all lines of sight are parallel and orthogonal to the plane, which eliminates the convergence of lines seen in perspective views.^[9] As a subset of parallel projection, it focuses on accurate scaling without angular distortion for engineering purposes.^[9] In multiview orthographic projection, an object is depicted through up to six principal views, each obtained by projecting the object onto a plane parallel to one of its principal faces: front, rear, top, bottom, left, and right.^[30] These views are typically arranged in a standardized layout, often using the "glass box" method where the object is imagined inside a transparent cube, and each face unfolds to form the drawing planes, with the front view serving as the central reference.^[31] While all six views provide a complete description, practical drawings commonly employ three—front, top, and right side—for sufficient detail, ensuring that corresponding features align vertically and horizontally across views.^[31] The arrangement of these views follows either first-angle or third-angle conventions, which differ in their spatial orientation relative to the object. In first-angle projection, adopted by ISO standards and common in Europe and Asia, the views are placed as if the object is in the first quadrant, with the top view below the front view, the right view to the left of the front, and so on, simulating the object projecting through the plane toward the observer.^[32] Conversely, third-angle projection, used in ANSI/ASME standards and prevalent in the United States, positions views as if observed from outside the object in the third quadrant, placing the top view above the front, the right view to the right, and others accordingly for intuitive alignment.^[32] Projection symbols, such as truncated cones, are included in drawings to indicate the method and prevent misinterpretation.^[32] This projection technique is a standard in mechanical engineering for creating dimensioned technical drawings, where precise measurements and tolerances are applied directly to the views for manufacturing and assembly.^[31] Hidden lines, represented as dashed lines, indicate internal features like holes or edges not visible in a given view, while shading or sectioning may be used to convey depth and material properties without altering the flat representation.^[31] Key properties of orthographic projection include the preservation of parallelism, where lines parallel in the object remain parallel in the projection, and the absence of perspective distortion, allowing for exact scaling of surfaces aligned with the projection plane.^[33] These characteristics make it ideal for applications requiring high accuracy in dimensioning, such as specifying manufacturing tolerances, as features parallel to the plane appear in their true proportions without foreshortening or convergence.^[33]

Oblique Projection

Oblique projection is a type of parallel projection in technical drawing where the projection lines are not perpendicular to the projection plane but instead meet it at an oblique angle, typically 45 degrees, allowing the representation of three-dimensional objects in two dimensions while preserving the true proportions of the front face. In this method, the front face of the object is depicted in its actual size and shape, with depth lines receding at the chosen angle to convey spatial extent without convergence.^[34] This approach balances simplicity with a partial illusion of depth, making it suitable for illustrations where one primary view needs to remain undistorted.^[35] Several variants of oblique projection exist, distinguished primarily by the scaling of the receding depth lines and the specific angles employed. The cavalier projection uses full-scale depth lines at a 45-degree angle, resulting in a straightforward but potentially exaggerated representation of depth.^[34] In contrast, the cabinet projection reduces the depth lines to half their true length, also at 45 degrees, to minimize distortion and provide a more balanced appearance.^[35] A related variant, the military projection—also known as plan oblique—employs 45-degree angles for both x-z and y-z axes while keeping the x-y plane perpendicular, often with full or foreshortened depth; it was historically used by French military artists in the 18th century for tactical sketches and maps, emphasizing clear floor plans.^[26]^[36] One key advantage of oblique projection is its ease of construction compared to perspective drawing, as it requires only basic geometric tools and avoids complex vanishing points, enabling quick sketches in fields like engineering and architecture. It also ensures that the front face remains true to scale, facilitating accurate measurement and assembly instructions in technical illustrations.^[34] This method finds practical use in woodworking plans and product illustrations, where undistorted frontal views aid in visualizing construction details without the need for multiple orthographic projections.^[37] In terms of properties, oblique projection renders circles and curves on the front face as true shapes, but those on receding planes appear as ellipses due to the angular skew, which can introduce foreshortening effects.^[35] While effective for rapid conceptualization, it may distort the proportions of rear and side faces, particularly in cavalier variants, making it less ideal for precise multi-angle representations when compared to axonometric methods that equalize all axes.

Axonometric Projection

Axonometric projection is a type of parallel projection technique used to represent three-dimensional objects on a two-dimensional plane, where the object's coordinate axes are inclined to the projection plane at equal or specified angles, ensuring that lines parallel in three-dimensional space remain parallel in the projection.^[24] This method allows for a single view that simultaneously displays three faces of an object, providing a pictorial representation without the convergence of lines to vanishing points, as all projectors are parallel.^[38] Unlike orthographic projection, which shows only one or two faces, axonometric projection rotates the object relative to the viewing plane to reveal depth in a balanced manner.^[39] The primary variants of axonometric projection are isometric, dimetric, and trimetric, distinguished by the scaling and angular relationships among the three axes.^[36] In isometric projection, all three axes are scaled equally (typically with a foreshortening factor of about 0.816) and projected at angles of approximately 120 degrees to each other, resulting in a symmetric view ideal for objects with cubic symmetry.^[40] Dimetric projection features two axes with identical scales and angles (such as 105 degrees each), while the third axis has a different scale and angle (e.g., 150 degrees), allowing for views where two dimensions are emphasized equally.^[41] Trimetric projection generalizes this further, with all three axes having distinct scales and angles, enabling highly customized orientations for specific visualization needs.^[42] Axonometric projections are widely used in technical illustrations, such as exploded assembly diagrams and piping isometrics in engineering, where preserving accurate measurements along the projected axes is essential for clarity and dimensioning.^[43] A key property is the uniform application of foreshortening to the axes based on their angles to the projection plane, which maintains proportional relationships without perspective distortion.^[38] To optimize the view, the object is often rotated so that its principal axes align with the projection directions, maximizing visibility and minimizing occlusion.^[40]

Mathematical Foundations

Formulas for Parallel Projection

Parallel projections map three-dimensional points onto a two-dimensional plane using parallel rays, resulting in an affine transformation that preserves parallelism and ratios along parallel lines. The general coordinate transformation for a point (x, y, z) in oblique parallel projection is given by

x' = x + z \cdot \cos \theta \cdot s, \quad y' = y + z \cdot \sin \theta \cdot s,

where \theta is the angle of the projectors relative to the projection plane (often 45°), and s is the foreshortening scale factor. In cavalier projection, s = 1, preserving full depth, while in cabinet projection, s = 0.5, reducing depth to half scale to minimize distortion.^[44]^[45] Orthographic projection, a special case of parallel projection with projectors perpendicular to the plane (\theta = 0^\circ), simply discards the z-coordinate, yielding (x', y') = (x, y) for the front view. Multiview orthographic projections require rotating the object to align principal faces with the projection plane before applying this. For the top view, a 90° rotation around the x-axis transforms (x, y, z) to (x, -z, y), represented by the matrix

\begin{pmatrix} 1 & 0 & 0 \\ 0 & 0 & -1 \\ 0 & 1 & 0 \end{pmatrix},

followed by dropping the new z-coordinate to obtain (x', y') = (x, -z). Similar rotations apply for side views, such as a 90° rotation around the y-axis for the right view.^[4]^[46] Axonometric projections extend orthographic methods by rotating the object to equalize or specify axis foreshortenings. For isometric projection, the object is first rotated 45° around the y-axis and then \arctan(1/\sqrt{2}) \approx 35.264^\circ around the x-axis, with uniform scaling by $1/\sqrt{2} along the projected axes to ensure equal lengths. The combined transformation matrix is the product of these rotation matrices followed by orthographic projection, often expressed in homogeneous coordinates as a 4×4 matrix:

M = \frac{1}{\sqrt{2}} \begin{pmatrix} 1 & 0 & 0 \\ 0 & \cos 35.264^\circ & \sin 35.264^\circ \\ 0 & -\sin 35.264^\circ & \cos 35.264^\circ \end{pmatrix} \begin{pmatrix} \cos 45^\circ & 0 & \sin 45^\circ \\ 0 & 1 & 0 \\ -\sin 45^\circ & 0 & \cos 45^\circ \end{pmatrix} \begin{pmatrix} 1 & 0 & 0 & 0 \\ 0 & 1 & 0 & 0 \\ 0 & 0 & 0 & 0 \\ 0 & 0 & 0 & 1 \end{pmatrix}.

Dimetric projections use two equal scales with unequal third, while trimetric uses three distinct scales, adjusting the rotation angles and factors accordingly (e.g., different \cos and \sin values for each axis).^[45]^[47] In implementation, parallel projections use vector notation for efficient edge rendering, where object edges are transformed linearly without depth division. As affine transformations, they inherently preserve collinearity and parallelism, ensuring straight lines remain straight and parallel lines do not converge, which is ideal for technical illustrations and CAD systems.^[44]^[4]

Formulas for Perspective Projection

Perspective projection, also known as central projection, models the way light rays from a 3D scene converge through a point (the camera center or viewpoint) onto a 2D image plane, mimicking the human eye or a pinhole camera. In a simple setup, consider the viewpoint at the origin (0,0,0) and the image plane at z = d, where d is the focal length or distance to the plane. A 3D point (X, Y, Z) in the scene, with Z > 0 facing the camera, projects to the image coordinates (x, y) on the plane z = d via similar triangles: the ray from the viewpoint through (X, Y, Z) intersects the plane at x = d \cdot \frac{X}{Z} and y = d \cdot \frac{Y}{Z}.^[48] This nonlinear division by depth Z ensures that objects farther away appear smaller, creating realistic depth cues.^[28] To handle this projection computationally, homogeneous coordinates are employed, representing 3D points as 4-vectors [X, Y, Z, 1] and 2D image points as [x, y, 1] after normalization. The perspective transformation can then be expressed as a 3x4 matrix multiplication followed by division by the third coordinate: if \mathbf{m} = P \begin{pmatrix} X \ Y \ Z \ 1 \end{pmatrix} = \begin{pmatrix} m_1 \ m_2 \ m_3 \end{pmatrix}, the projected point is (x, y) = (m_1 / m_3, m_2 / m_3).^[49] For the canonical case with viewpoint at (0,0,0) and plane at z = d, this simplifies to the dehomogenized form [X/Z \cdot d, Y/Z \cdot d, 1], aligning with the earlier equations.^[50] A key feature of perspective projection is the convergence of parallel lines to vanishing points, which occur because parallel lines in 3D correspond to rays that meet at a point at infinity. For a set of parallel lines in direction vector (a, b, c), their projections intersect at the vanishing point (x_v, y_v) = \left( d \cdot \frac{a}{c}, d \cdot \frac{b}{c} \right) on the plane z = d, assuming c \neq 0 (i.e., the direction is not parallel to the image plane).^[51] If the direction is oblique or parallel to the plane (c = 0), multiple vanishing points may arise, one for each set of parallels not aligned with the optical axis, as seen in multi-point perspective drawings.^[28] When objects lie at roughly similar depths relative to the camera (small depth variation compared to average distance), the full perspective model can be approximated by weak perspective projection to simplify computations, particularly in pose estimation tasks. Here, the projection is treated as a scaled orthographic projection: let Z_0 be the average depth, and s = d / Z_0 the uniform scale factor; then a point (X, Y, Z) projects approximately to (x, y) = (s X + u_0, s Y + v_0), where the depth Z is ignored in the lateral coordinates but the scale accounts for average distance. This is derived from linearizing the division: x \approx s X \left(1 - \frac{Z - Z_0}{Z_0}\right) for small deviations, though the basic form omits the higher-order term for approximation.^[49] This model reduces nonlinearity while retaining some depth scaling, making it suitable for scenarios like human pose estimation from video where depth variations are limited.^[52] In computer vision, the full perspective projection is formalized through the pinhole camera model using a 3x4 projection matrix P that combines intrinsic and extrinsic parameters. The intrinsics are captured by the 3x3 upper-triangular matrix K = \begin{pmatrix} f_x & s & u_0 \ 0 & f_y & v_0 \ 0 & 0 & 1 \end{pmatrix}, where f_x, f_y are focal lengths in pixel units, s is skew, and (u_0, v_0) is the principal point; the extrinsics are the rotation R (3x3) and translation t = -R C (3x1), with C the camera center. Thus, P = K [R | t], and the projection is \begin{pmatrix} x \ y \ w \end{pmatrix} = P \begin{pmatrix} X \ Y \ Z \ 1 \end{pmatrix}, followed by (x/w, y/w).^[53] This derives from the pinhole geometry: first, transform world coordinates to camera frame via [R | t], yielding (X_c, Y_c, Z_c) = R (X_w - C); then apply perspective as x = f_x X_c / Z_c + u_0, y = f_y Y_c / Z_c + v_0, which in matrix form incorporates the division via homogeneous coordinates.^[49]

Applications and Limitations

Applications in Various Fields

In technical drawing and engineering, orthographic projection serves as the standard for creating precise blueprints and multiview representations of three-dimensional objects, enabling accurate dimensioning and fabrication. This method adheres to ASME Y14.3 standards, which define requirements for orthographic and pictorial views on engineering drawings to ensure consistency in mechanical design and manufacturing processes.^[54] Axonometric projection, particularly isometric variants, is widely employed in assembly instructions for manufacturing, providing a clear, undistorted pictorial overview of component relationships without perspective convergence, as seen in exploded views for machinery and product assembly. In art and architecture, perspective projection revolutionized visual representation during the Renaissance, allowing artists to depict spatial depth realistically in paintings and frescoes. Filippo Brunelleschi's early 1400s experiments with linear perspective laid the groundwork, influencing works like Masaccio's The Holy Trinity fresco, where converging lines create an illusion of three-dimensional space on flat surfaces.^[55] Oblique projection finds practical use in architectural interior sketches, where the plan is rotated at 45 degrees to illustrate room layouts and furniture placement while maintaining true dimensions on horizontal surfaces, facilitating quick conceptual designs.^[56] Computer graphics and gaming rely heavily on perspective projection for rendering immersive three-dimensional scenes, as implemented in APIs like OpenGL, where it simulates human vision by scaling objects based on distance from the viewpoint.^[57] Parallel projections, such as orthographic views, are commonly applied to user interface elements like minimaps in video games, ensuring consistent scaling and navigation aids without depth distortion, as exemplified in titles using top-down overviews for strategic gameplay.^[58] In medical imaging, orthographic projection underpins the reconstruction of computed tomography (CT) scans, where parallel slices generate cross-sectional views of anatomy, enabling precise volumetric analysis without perspective artifacts.^[59] Cartography utilizes various azimuthal projections, including perspective types like gnomonic and parallel types like orthographic, to represent polar regions or hemispheric data accurately, preserving directions and distances for applications like aviation charts and Antarctic exploration maps.^[60] In virtual reality (VR) and augmented reality (AR), real-time perspective projection enhances immersion by dynamically rendering environments that align with user head movements, with post-2010 advancements in head-mounted displays enabling low-latency 3D overlays on the real world.^[61] Modern software tools integrate these projections seamlessly; AutoCAD employs parallel orthographic views for engineering drafting and precise 2D/3D modeling in construction projects.^[62] Blender supports both perspective and parallel projections through configurable cameras, facilitating artistic rendering and animation workflows. Recent integrations with artificial intelligence in visualization software, such as AI-driven segmentation in Amira and Avizo, automate the generation and projection of 3D models from imaging data, streamlining analysis in fields like materials science.^[63]^[64]

Limitations and Comparisons

Parallel projections, while preserving parallelism and measurements along projection directions, suffer from a lack of depth realism, causing objects to appear uniformly sized regardless of their distance from the viewer, which results in a flat, unrealistic representation.^[65] In orthographic projections specifically, the absence of depth cues can obscure overlaps between objects, making it difficult to discern spatial relationships that would be evident in more realistic views. Oblique and axonometric projections within the parallel family introduce distortions in non-axial measurements, where receding lines appear foreshortened or elongated, potentially misrepresenting object proportions unless compensated for during creation.^[67] Perspective projections, in contrast, introduce depth distortion particularly for objects near the viewer, where proximity causes disproportionate magnification and nonlinear scaling that alters relative sizes and shapes.^[69] They are computationally more intensive than parallel methods due to the need for division operations and handling of converging rays, increasing processing demands in rendering pipelines.^[69] Additionally, the presence of vanishing points, where parallel lines converge, complicates uniform scaling and alignment in multi-object scenes, as each set of parallels requires individual handling to maintain consistency.^[3] Comparisons between projection types highlight key trade-offs: parallel projections prioritize measurement accuracy and simplicity, ideal for technical drawings where exact dimensions are critical, whereas perspective projections emphasize visual realism at the expense of proportional fidelity.^[70] Orthographic projections excel in multiview precision for engineering analysis, providing unambiguous planar representations, while axonometric projections offer intuitive single-view overviews but with added foreshortening that aids spatial comprehension in design reviews.^[1] Selection depends on context; engineering and CAD applications favor parallel methods for their distortion-free metrics, whereas art, architecture, and CGI prioritize perspective for immersive, human-like perception.^[71] To address these limitations, hybrid approaches like weak perspective projection approximate full perspective by scaling an orthographic base, mitigating nonlinearity for stable computations in computer vision tasks while introducing minimal inaccuracy for distant objects.^[72]^[73] Software corrections, such as anti-aliasing filters, counteract distortions like jagged edges in digital projections, a longstanding issue in CGI where sampling artifacts can appear in rendered perspective views.

References

[1]
Introduction to Computer Graphics, Section 3.3 -- Projection and ...
The coordinate system on this cube is referred to as clip coordinates. The transformation from eye coordinates to clip coordinates is called the projection ...
[2]
Projections in Computer Graphics - GeeksforGeeks
Dec 6, 2022 · Projection converts a 3D object to 2D, representing it on a plane. Types include parallel and perspective projections. Parallel lines extend to ...
[3]
The Perspective and Orthographic Projection Matrix - Scratchapixel
The aim of the orthographic projection matrix is to remap all coordinates within a specific 3D space bounding box to the canonical viewing volume.
[4]
The Perspective and Orthographic Projection Matrix - Scratchapixel
Projection matrices are 4x4 matrices that transform 3D points to 2D coordinates on a canvas, projecting 3D objects onto the screen.Exploring Projection... · Understanding Their... · Orthographic Vs. Perspective...<|control11|><|separator|>
[5]
[PDF] Projections - Washington
Projections transform points in n-space to m-space, where m < n. In 3D, we map points from 3-space to the projection plane (PP) along projectors emanating from ...
[6]
[PDF] 3D Viewing
Viewing and Projection. • Our eyes collapse 3-D world to 2-D retinal image. (brain then has to reconstruct 3D). • In CG, this process occurs by projection.
[7]
Intro to Computer Graphics: Viewpoints and Projections
Projection is 'formed' on the view plane (planar geometric projection) rays (projectors) projected from the center of projection pass through each point of the ...
[8]
Computer Graphics - 3D Projection and Visualization
3D to 2D mapping is called projection, involving camera, projection, and viewport transformations. Types include orthographic and perspective projections.
[9]
Projections and Views | Engineering Design - McGill University
A three-dimensional object can be represented in a single plane, such as on a sheet of paper, using projecting lines and planes.
[10]
Geometry in Art & Architecture Unit 11 - Dartmouth Mathematics
Perspective is an example of the geometric operation of projection and section where projection lines from the outline of an object to the eye are sectioned or ...
[11]
[PDF] 3D Viewing - Department of Computer Science
Think of 3D viewing as taking a photo: – Select Projection. – Specify viewing parameters. – Perform trivial accept/reject test in 3D.<|control11|><|separator|>
[12]
[PDF] Projections - Washington
Parallel projection pros and cons: – Less realistic looking. + Good for exact measurements. + Parallel lines remain parallel. – Angles not (in general) ...Missing: engineering | Show results with:engineering
[13]
[PDF] Axonometric Projections - andrew.cmu.ed
Axonometric projections have the advantage that they give a pictorial view of the object, yet dimensions are measurable. Manually, axonometric views can be ...
[14]
[PDF] Adam Rilling Math 308 Project Projections Projections hold an ...
Whatever matrix is chosen, linear algebra is the key to the transformation. These engineering tools use parallel as well as perspective projection to display ...
[15]
Mathematics of Perspective Drawing - University of Utah Math Dept.
Perspective drawing involves mapping points in 3D space to a drawing plane using perspective transformations, which are a composition of rigid motion and ...<|control11|><|separator|>
[16]
[PDF] 3D Viewing and Projection Reading: Chapter 5 Review
Projection. Projection Plane. Viewer. Scene. Image. • The main types of projection in computer graphics are: • Parallel (orthographic) projection. • Perspective ...
[17]
[PDF] The Geometry of Perspective Projection
- It is the projection of a 3D object onto a plane by a set of parallel rays orthogonal to the image plane. - Parallel lines project to parallel lines. - Size ...Missing: principles | Show results with:principles
[18]
[PDF] Computer Vision: Projection - Washington
Variants of orthographic projection. Scaled orthographic. • Also called ―weak perspective‖. Affine projection. • Also called ―paraperspective‖. Projection ...
[19]
Introduction to Orthographic Projection - SDC Publications
What is an Orthographic Projection? An Orthographic projection is a 2-D representation of a 3-D object. The 2-D drawing represents different sides of an object.
[20]
[PDF] orthographic projection
Orthographic projection is a way to represent a 3D object on a piece of paper: The object becomes 2D. No single drawing is able to show what an object truly ...
[21]
First vs Third Angle – Orthographic Views | GD&T Basics
Orthographic views allow us to represent a 3D object in 2D on a drawing. Orthographic views can show us an object viewed from each direction.
[22]
Orthographic Projection – Engineering Graphics and Design
Any dimensions or features of the object which are parallel to the projection plane will be shown in true size and shape (not distorted). Any features lying on ...Missing: properties no
[23]
Architecture Drawing Projections
Jun 29, 2023 · Cavalier projection is when one axis is horizontal (x), one axis is vertical (y) and the third axis (z) is drawn at an angle of 30°, 45° or 60° ...
[24]
Oblique Projection in Computer Graphics - Tutorials Point
Oblique projection provides a way to represent objects by displaying one face of the object in true size while the other faces are skewed.
[25]
Graphical Projection | Encyclopedia MDPI
Nov 7, 2022 · Unlike cavalier projection, where the third axis keeps its length, with cabinet projection the length of the receding lines is cut in half.<|control11|><|separator|>
[26]
Oblique Drawings in a Flash | Blog | Videos | Plans | How To
Jul 13, 2010 · I'm not a whiz at the drafting table, but here's my secret for making oblique perspective drawings without a T-square and triangles.
[27]
Axonometric projections - a technical overview - CompuPhase
May 28, 2019 · Axonometry is a drawing technique where the orthogonal x-, y- and z-axes of the (3-dimensional) world space are projected to (non-orthogonal) axes on paper.Missing: trimetric scholarly
[28]
[PDF] CHAPTER 4 - UNM CS
Finally, we derive the projection matrices that describe the most important parallel and perspective views and investigate how to carry out these projections ...
[29]
[PDF] pictorials - De Anza College
Engineers and designers use pictorial sketches to refine and communicate 3D designs before they are formally drawn or modeled. The choice of pictorial drawing ...
[30]
[PDF] preliminary draft, October 25, 2002 1:43 preliminary draft ... - LORIA
The projections in the second row are oblique projections, where the projection plane is not orthogonal to the projection direction. The trimetric projection is ...
[31]
[PDF] Perspective Projection and Arbitrary Viewing
There are two primary categories of parallel projections: orthographic and oblique. Orthographic parallel projection assumes that the plane of projection is ...Missing: engineering | Show results with:engineering<|separator|>
[32]
[PDF] Produce Proportional Views of Axonometric Projections
The three basic types of axonometric projection are isometric, diametric, and trimetric. A. Isometric projection and isometric drawing. 1. Isometric drawing ...Missing: variants scholarly
[33]
[PDF] 5.8.3 Oblique Projections - UNM CS
xp = − x z , yp = − y z . We have shown that we can apply a transformation N to points, and, after an orthogonal projection, we obtain the same result as we ...
[34]
Parallel Projection - Graphics Programming - Part 6 - Chapter 2
The projection rays are parallel, i.e., the projection center is located at infinity; The formula for mapping a 3D point \mathbf{P}=(p_x,p_y,p_z)^\top to a ...
[35]
OpenGL Projection Matrix - songho.ca
GL_PROJECTION matrix is used for this projection transformation. First, it transforms all vertex data from the eye coordinates to the clip coordinates.Perspective Matrix With... · Using Vertical Fov · Using Horizontal Fov<|separator|>
[36]
3d isometric projection matrix? - Game Development Stack Exchange
Mar 12, 2019 · An isometric projection matrix is an orthographic matrix with specific camera placement, often created with built-in methods in modern matrix ...How to convert 3D coordinates to 2D Isometric Coordinates?Why when I do an isometric projection do I get a "dimmetric" view?More results from gamedev.stackexchange.com
[37]
[PDF] 5.4.1 Perspective Projections - UNM CS
and xp = x z/d . Using the side view, we obtain a similar result for yp, yp = y z/d . These equations are nonlinear. The division by z describes nonuniform ...
[38]
[PDF] CS231A Course Notes 1: Camera Models
In this section, we will discuss the details of the parameters we must account for when modeling the projection from 3D space to the digital images we know. All ...
[39]
[PDF] Perspective Projection
Let n = (a, b, c, d) be the plane vector of a viewplane described by the general equation ax + by + cz + d = 0, and v be a point not on the viewplane. The ...Missing: axonometric | Show results with:axonometric
[40]
[PDF] Equation of Perspective Projection Projective Geometry Conversion
In the projective space, parallel lines meet at a point at infinity. • The vanishing point is the perspective projection of that point at infinity, resulting.
[41]
[PDF] Cameras | CS5670: Computer Vision
Perspective projection. Page 28. Variants of orthographic projection. • Scaled orthographic. – Also called “weak perspective”. • Affine projection. – Also ...
[42]
[PDF] Multiple View Geometry in Computer Vision, Second Edition
PART 0: The Background: Projective Geometry, Transformations and Esti- mation. 23. Outline. 24. 2. Projective Geometry and Transformations of 2D.
[43]
Y14.3 - Orthographic and Pictorial Views - ASME
In stockThis Standard establishes the requirements for creating orthographic, and pictorial views on engineering drawing graphic sheets and in models.Missing: projection | Show results with:projection
[44]
PROJECTION METHODS ON ENGINEERING DRAWINGS
Engineering projection methods enables mechanical desing engineer to represent 3D object in 2D. Learn more about projection methods.Pictorial drawing projections... · Perspective projection · Orthographic drawing...
[45]
Gaining Perspective - Nelson Atkins
In the early 1400s, the Italian architect Filippo Brunelleschi (1377–1446) reintroduced a means of rendering the recession of space, called linear perspective.
[46]
Plan Oblique Illustration: Part 1 | Visualizing Architecture
Jul 29, 2012 · This first part will describe how to generate a true plan oblique view. Later posts will describe the methods I used to add color and texture.
[47]
3D Graphics with OpenGL - The Basic Theory
The projection with view frustum is known as perspective projection, where objects nearer to the COP (Center of Projection) appear larger than objects further ...
[48]
Game developer's guide to graphical projections (with ... - Medium
Dec 2, 2017 · As one of the parallel projections, objects drawn with multiview projections retain size no matter where in space they are positioned.
[49]
Orthographically projected CT and depth images. Left image shows ...
Orthographically projected CT and depth images. Left image shows a CT abdomen scan in a larger scale. Middle image shows a depth image rendered in color.
[50]
Azimuthal Projection: Orthographic, Stereographic and Gnomonic
The azimuthal projection plots the surface of the Earth using a flat plane. Imagine light rays radiating from a source following straight lines.
[51]
Augmented reality and virtual reality displays: emerging ... - Nature
Oct 25, 2021 · VR embraces a total immersive experience, while AR promotes the interaction between user, digital contents, and real world, therefore displaying ...
[52]
3D CAD Software for Drawing & Modeling - Autodesk
Learn how Autodesk's 3D CAD design software helps users design ideas, visualize concepts, and simulate how designs will perform in the real world.
[53]
Blender - The Free and Open Source 3D Creation Software ...
Blender includes production ready camera and object tracking. Allowing you to import raw footage, track the footage, mask areas and see the camera movements ...Download · Blender Foundation · Blender Institute · Blender Store
[54]
Artificial Intelligence for 3D Visualization and Analysis Software
Accelerate your imaging data processing, segmentation, and classification using Machine and Deep Learning in Amira, Avizo, and PerGeos Software.
[55]
[PDF] Tyler Pringle MATH 400 Presentation 2
A limitation of this kind of projection (and of parallel projections in general) is that objects do not appear larger as they extend closer to the camera, ...
[56]
Orthographic Drawing: Plans & Sections - GSAPP SMORGASBORD
Dec 14, 2021 · They show the flat planes of a space as seen straight on, without representing depth or volume. This limits what they can show, but makes what ...
[57]
[PDF] OBLIQUE AND AXONOMETRIC PROJECTIONS are useful
In an oblique projection, the depth of the object in the angled view can be distorted. So you must compensate for this if you want to represent the object in a ...
[58]
3.4: Oblique projections - Engineering LibreTexts
Sep 13, 2025 · In oblique drawings, surfaces that are aligned with the front plane appear in their true shape and size, as they would appear in the front view.Missing: variants military
[59]
Perspective Projection in Computer Graphics - Tutorials Point
Limitations of Perspective Projection · Complexity − Compared to orthographic or oblique projection, perspective projection is mathematically more complex.
[60]
Difference between Parallel and Perspective Projection in Computer ...
May 24, 2020 · Parallel projection can give the accurate view of object. Perspective projection cannot give the accurate view of object.
[61]
AutoCAD 2022 for Mac Help | About Parallel and Perspective Views
The difference between perspective views and parallel projections is that perspective views require a distance between a theoretical camera and target point. ...
[62]
Weak Perspective Projection
As can be seen in the entry perspective camera, the perspective projection is a nonlinear mapping. This makes many vision problems difficult to solve, and.Missing: advantages | Show results with:advantages
[63]
What can be determined from a full and a weak perspective image?
The weak perspective approximation leads to more stable estimation algorithms, although at the cost of a small amount of modeling inaccuracy. Previous work in ...Missing: advantages | Show results with:advantages
[64]
[PDF] Antialiasing for Automultiscopic 3D Displays - UMD Computer Science
We then present an anti-aliasing technique for automultiscopic displays by combining a reconstruction and a display prefilter. Next, we show how to ...Missing: CGI | Show results with:CGI