Mixed reality (MR) is a class of interactive computing in which digital objects, information or environments are spatially integrated with the user's physical surroundings so that physical and virtual elements can be perceived and, to varying degrees, interact with one another. The term is often used for experiences that go beyond simply displaying graphics over a camera image or transparent view of the real world: virtual content can appear anchored to particular locations, surfaces or objects and can respond to the user's movement and the surrounding environment.
MR therefore depends on some combination of environmental sensing, spatial mapping, positional tracking, rendering and human-computer interaction. Depending on the technical definition being used, MR can encompass experiences ranging from sophisticated augmented reality to immersive systems in which virtual and physical elements coexist. It is consequently better understood as a spectrum or category of spatial interaction rather than as one single display technology.
The underlying principle of mixed reality is spatial registration: the computer must maintain a sufficiently accurate relationship between virtual content and the physical environment so that digital objects appear to occupy stable positions. Sensors can gather information about head position, device movement, surfaces, depth, lighting and other environmental characteristics, while software constructs or maintains a representation of the user's surroundings. A rendering system then generates the appropriate perspective of virtual content as the user moves.
If the system recognises a physical table, for example, a virtual object may be positioned on its surface and continue to appear attached to that location as the user walks around it. More advanced systems can also use occlusion information so that a physical object can visually obscure a virtual object, producing a stronger impression that both occupy the same space.
Mixed reality is closely related to augmented reality (AR), virtual reality (VR) and extended reality (XR), but the terms are not interchangeable in every technical context. AR generally adds digital information to perception of the real world, whereas VR replaces or substantially excludes the physical environment with a computer-generated environment.
MR is commonly used for experiences in which virtual and physical elements are more deeply spatially integrated, although industry terminology varies and some products marketed as MR are technically advanced AR systems. XR is the broader umbrella term normally used for technologies that combine or extend physical and digital experiences, potentially including AR, VR and MR. Because there is no universally enforced boundary between these categories, a particular application may be described differently by researchers, manufacturers and developers while using substantially similar underlying technologies.
A mixed-reality system can involve several hardware components and software layers working together. Cameras and depth sensors may observe the environment; inertial measurement units containing accelerometers and gyroscopes measure movement; positioning systems estimate the device's pose; computer-vision algorithms detect features, planes and objects; and spatial-mapping systems construct representations of the surrounding environment. A display then presents rendered virtual imagery from a viewpoint corresponding to the user's position and orientation.
Interaction can involve hand tracking, controllers, voice commands, gaze, gestures or physical objects. Some systems use passthrough video, in which cameras capture the real environment and displays present that camera view alongside computer-generated content. Others use optical see-through displays that allow the wearer to look directly at the physical environment while projected or emitted light forms virtual imagery within the visual field.
Mixed reality has applications in training, engineering, design, healthcare, education, manufacturing, maintenance, architecture, retail, entertainment and collaborative work. An engineer can inspect a three-dimensional model at the scale of a physical machine, while a technician can receive spatially positioned instructions alongside equipment. Medical education can use volumetric visualisations of anatomy, and industrial training can simulate procedures while retaining awareness of a real training environment. Designers can place virtual prototypes in physical spaces before manufacturing them.
In education, spatial visualisations can make abstract three-dimensional relationships directly manipulable. The practical value depends on factors such as tracking accuracy, display quality, field of view, latency, environmental understanding, interaction design and whether the digital content genuinely benefits from spatial presentation rather than being more efficiently delivered through a conventional screen.
A major technical challenge is maintaining low-latency and accurate alignment between the user's physical environment, their movements and rendered virtual content. Tracking errors can cause virtual objects to drift, jump or appear incorrectly positioned. Sensor noise, reflective surfaces, poor lighting, visually repetitive environments and rapidly changing scenes can complicate environmental reconstruction. Occlusion is particularly difficult because the system needs to determine which physical surfaces should appear in front of virtual objects.
Realistic lighting is another challenge: virtual objects can look visually detached if their brightness, shadows and reflections do not correspond sufficiently to the physical environment. Computational requirements also increase when systems perform continuous spatial mapping, object recognition, rendering and interaction simultaneously. Privacy is an additional consideration because spatial systems can collect detailed information about rooms, physical objects, movement patterns and potentially people present in the environment.
Mixed reality does not necessarily mean that virtual objects physically exist or that a computer has perfect knowledge of the environment. A virtual object can be spatially anchored and interactable without having any physical presence, and environmental understanding is generally probabilistic rather than exact. Likewise, a headset that provides a live camera view with digital overlays is not automatically capable of all forms of mixed-reality interaction.
The distinction often concerns the sophistication and purpose of spatial integration rather than merely whether graphics are visible over reality. Another misconception is that MR requires a particular headset category; in practice, MR experiences can be delivered through different optical and video-based display architectures, provided the necessary sensing, rendering and interaction capabilities are available.
The modern concept of mixed reality developed alongside advances in computer graphics, computer vision, wearable computing and human-computer interaction. The term became particularly influential after research into mixed-reality environments described a continuum between entirely physical and entirely virtual experiences, providing terminology for intermediate combinations of real and synthetic content. Subsequent developments in mobile processors, depth sensing, simultaneous localisation and mapping, compact displays and machine learning made spatially aware consumer and industrial systems increasingly practical.
Contemporary MR therefore represents the convergence of several previously separate technologies rather than a single invention. Its defining characteristic is the integration of digital and physical spatial information into an interactive experience, with the degree of integration determined by the capabilities of the particular hardware and software system.
