Best 3D Scanner Apps in 2026

Summary: Best 3D Scanner Apps in 2026

The best 3D scanner apps in 2026 are the ones whose capture method, processing model, and export path match the job. If you need room capture on a LiDAR iPhone, the shortlist looks different from what you would choose for textured collectibles, 3D printing reference meshes, or fast visual sharing.

There is no single winner because mobile scanning apps do not all produce the same kind of output. Some are mainly mesh tools for OBJ, STL, GLB, or USDZ workflows. Some can also output point-cloud-style data. Some now produce Gaussian splats, which are useful visual deliverables for view-dependent rendering but are not the same as an editable polygon mesh. That distinction matters more than any global rank, because a room mesh for remodeling, a textured asset for a game engine, and a splat for quick visual capture solve different problems. [S23]

Best apps by job:

• Rooms and spaces on LiDAR iPhone: Polycam or 3D Scanner App.
• Small textured objects: RealityScan Mobile or KIRI Engine.
• Featureless or low-texture objects where officially supported: KIRI Engine.
• Textured game and AR assets: RealityScan Mobile or KIRI Engine.
• 3D printing reference meshes: Polycam, KIRI Engine, or 3D Scanner App, followed by cleanup.
• Fast visual sharing with splats: Scaniverse or KIRI Engine.
• iPhone-specific mixed workflows: Polycam for spaces, RealityScan or KIRI for objects.
• Android-first workflows: KIRI Engine, RealityScan Mobile, and Scaniverse.

How we ranked these apps

This shortlist ranks by job plus deliverable class, not by a single universal score. A mesh, a point cloud, and a Gaussian splat are not interchangeable just because they all come from a phone scan. Meshes fit editing, slicers, and many asset pipelines. Point-cloud-style outputs fit some inspection, documentation, and conversion workflows. Splats are mainly for visual scene representation and novel-view rendering. Gaussian splats are evaluated as visual deliverables, not as STL/OBJ mesh competitors. [S23]

We judged each app using six criteria: primary deliverable; capture inputs and hardware dependency; processing model; export and workflow fit; reliability signals from official documentation; and scope limits such as rooms, small objects, or difficult surfaces. An app scores well when its official docs clearly state what hardware it needs, where processing happens, what formats it exports by mode or tier, and what it is actually meant to capture. It scores lower when those details are vague, contradictory, or easy to confuse across iPhone, Android, LiDAR, photogrammetry, and splat modes.

We did not rank by price, subscription value, or app-store stars, because those are unstable and often unrelated to whether the output suits the downstream workflow. We also did not flatten splats, meshes, and point clouds into one artificial leaderboard.

Device & capture-method matrix

LiDAR scanning uses active depth sensing from a rear scanner and is most useful for rooms, larger objects, and spatial context on supported Apple devices. Apple’s ARKit documentation says scene reconstruction requires a device with a LiDAR Scanner. [S07] ARKit depth data is not just a generic image effect: the depthMap stores distance in meters, and the confidenceMap expresses ARKit’s confidence in those depth values. [S08] TrueDepth is Apple’s front depth stack, useful for close-range capture examples such as Heges, but it is not a replacement for rear LiDAR room capture. Heges’ listing explicitly distinguishes selfie TrueDepth capture from rear LiDAR capture. [S25] Photogrammetry uses overlapping RGB photos or video to reconstruct a textured mesh. Gaussian splats use many 3D Gaussian primitives for visual scene representation and real-time radiance-field rendering rather than a conventional editable mesh. [S23]

On Android, depth support is broad but not equivalent to Apple’s rear LiDAR path. Google says that, as of May 2026, over 88% of active devices support ARCore’s Depth API. [S09] But Google also says the Depth API uses depth-from-motion and merges data from available sensors such as ToF, with the most accurate results typically coming at about 0.5 m to 5 m from the scene. [S10]

Table note: Apple’s scene reconstruction path is LiDAR-gated, ARKit depth values are measured in meters with confidence data, and ARCore Depth is a depth-from-motion system that can fuse ToF input rather than a direct Apple-style LiDAR equivalent. [S07] [S08] [S09] [S10]

Best 3D scanner apps in 2026: shortlist comparison

Shortlist grouped by job plus deliverable, with the output class called out explicitly so a splat workflow is never mistaken for a print-oriented mesh workflow.

Table note: Mode, tier, platform, and export path matter as much as the app name. Polycam, KIRI Engine, and Scaniverse each change behavior by mode; RealityScan is comparatively focused around a photogrammetry-to-Sketchfab pipeline. [S11] [S12] [S15] [S16] [S18] [S19] [S20] [S21]

App-by-app notes

These are workflow notes, not full reviews. The goal is to show where each app fits, what it really outputs, and what its documentation makes clear or unclear.

Polycam

According to Polycam’s help center, its processing model changes sharply by mode: LiDAR, Space, and Floorplan scans are processed locally on the device, while Object Mode, 360 captures, and Splats require cloud processing. [S11] That makes Polycam one of the clearest examples of why “does it work offline?” is not a yes-or-no question. If you mainly scan rooms on a LiDAR iPhone, Polycam is a strong fit because capture and first-pass processing stay local. If you mainly use object photogrammetry or splat workflows, internet access becomes part of the pipeline.

Polycam’s export documentation is also explicit that formats depend on both capture type and subscription tier, with Free limited to GLTF exports, Basic listing 12 export formats, and higher tiers adding point-cloud outputs. [S12] It is also less symmetrical across platforms than many roundups imply. Polycam’s current support page says Android Space Mode is not available, which rules out a simple iPhone-versus-Android parity claim for room scanning. [S13]

KIRI Engine

KIRI Engine stands out because its official docs separate the product into distinct modes instead of presenting one capture path as universal. Its export-formats page lists broad outputs for Photo Scan and Featureless Object Scan, including OBJ, FBX, STL, GLB, GLTF, USDZ, PLY, and XYZ, while LiDAR Scan is narrower at OBJ and USDZ. [S14] That makes KIRI especially practical for users who know their downstream toolchain first and want to work backward from the required handoff format.

KIRI is also a good example of why processing has to be labeled by mode. Its 3D Gaussian Splatting guide says processing runs in the cloud. [S15] Its AI-Enhanced LiDAR article says standard LiDAR processing can be on-device, while the AI-enhanced variant uploads for cloud processing. [S16] So KIRI is attractive when you want one app that spans photogrammetry, difficult low-texture objects, LiDAR on supported Apple hardware, and splats, but its behavior should not be reduced to simply local or simply cloud.

RealityScan Mobile

RealityScan Mobile is the most clearly photogrammetry-first app in this list. Its official page says it runs on iOS 16.1 or later and Android 7/API 24 or later on ARCore-supported devices. [S17] That makes it broadly accessible across both platforms without requiring LiDAR. It is a strong fit for textured object capture, especially if the end goal is a shareable asset rather than a room mesh.

Its workflow is also unusually specific. Epic’s documentation says local project files store the model as OBJ on iOS and GLB on Android. [S18] Unreal’s RealityScan page says the app is free to download and that models exported via Sketchfab can be downloaded in FBX, USDZ, and GLB or glTF formats. [S19] In practice, that makes RealityScan a strong candidate for textured assets headed toward Sketchfab, Unreal-adjacent workflows, or general asset libraries, but a less natural choice for LiDAR room capture or direct splat workflows.

Scaniverse

Scaniverse now spans more than the older “free splat app” reputation many readers still remember. Niantic Spatial’s Scaniverse Quickstart for Sites says each scan supports up to five minutes of recording time and up to 500 square meters of coverage, then can be uploaded for processing into mesh or Gaussian splat outputs with exports including .ply, .spz, .glb, .fbx, and .usdz. [S20] Niantic’s SPZ repository further defines .spz as a compressed 3D Gaussian splat format. [S22]

The main caution is that Scaniverse documentation is split across more than one workflow surface. The Sites docs describe upload-based asset generation, while the current Android Google Play listing says the app works “on your phone, and now in the cloud,” and lists SPZ, PLY, GLB, and FBX exports. [S21] At publication check, the safest interpretation is that Scaniverse processing is mode-dependent, not uniform. That makes it a strong recommendation for visual scene capture, splat sharing, and site-style scanning, but you should verify the exact mode you plan to use instead of assuming every Scaniverse capture path behaves the same way.

3D Scanner App (Laan Labs)

3D Scanner App remains one of the simpler LiDAR-oriented entries for iPhone and iPad users who want a direct scan-to-file path without much ecosystem framing. The current public-facing site lists OBJ, STL, PLY, USDZ, and GLTF exports. [S24] That is enough to cover basic mesh transfer to Blender, slicers, AR viewers, and common asset tools.

The reason it belongs on this list is not breadth across every capture technology, but focus. If your main need is a practical LiDAR utility on Apple hardware and you are comfortable doing cleanup elsewhere, 3D Scanner App is easy to place. For this article, the current official site is treated as the authoritative export list rather than older secondary pages with broader or conflicting claims. [S24]

Heges

Heges is useful less as a universal recommendation and more as a clear TrueDepth-versus-LiDAR example. Its App Store listing says it can scan with the selfie TrueDepth camera and the rear LiDAR sensor, and export PLY, OBJ, USDZ, STL, and GLB. [S25] That makes it relevant when you want to understand how close-range front-depth capture differs from rear-depth room capture on Apple devices.

The App Store listing contains precision language; we’re not treating it as validated accuracy. [S25] That caution matters because Heges’ listing includes marketing-style claims about precision settings and relative sensor quality, but this article separates capability claims from independently validated performance. So Heges is best read here as a useful iPhone-only example of the two Apple depth paths rather than a metrology-grade benchmark.

Best 3D scanner app for iPhone in 2026

On iPhone, the first split is simple: LiDAR-capable versus non-LiDAR. Apple’s scene-reconstruction path requires rear LiDAR hardware, so room scanning recommendations should start there rather than treating every iPhone as equivalent. [S07] ARKit depth values are measured in meters and accompanied by a confidence map, which is why LiDAR-based apps can make better-informed decisions about uncertain pixels. [S08] Polycam’s current support page also makes the practical hardware boundary explicit by listing LiDAR iPhones from the iPhone 12 Pro family onward for Space Mode. [S13] Non-LiDAR iPhones are still useful, but they are usually better matched to photogrammetry-heavy object capture. TrueDepth exists on some iPhones as a front-facing option, yet it should be treated as a separate close-range path rather than a substitute for rear LiDAR room scanning. [S25]

If you have a LiDAR iPhone and care about rooms or spaces, Polycam and 3D Scanner App are the most straightforward fits. If you care about textured small objects, RealityScan Mobile and KIRI Engine are stronger choices because they are built around RGB reconstruction rather than room-first depth capture. If you want visual splats or site-style scene sharing, Scaniverse and KIRI Engine make more sense than trying to force a room-mesh app into a splat job. If you want 3D printing reference geometry, Polycam, KIRI, and 3D Scanner App can all get you to STL-capable handoff paths, but you should still expect cleanup and dimensional validation before printing. [S11] [S12] [S14] [S16] [S20]

• Choose Polycam if you have a LiDAR iPhone and want room capture with on-device Space or Floorplan processing. [S11] [S13]
• Choose RealityScan Mobile if your priority is textured object capture and a Sketchfab-centered export path. [S17] [S18] [S19]
• Choose KIRI Engine if you want one app that spans photogrammetry, featureless-object support, LiDAR modes, and splats with broad export coverage. [S14] [S15] [S16]
• Choose Scaniverse if you want fast visual scene sharing and are comfortable with mesh-or-splat outputs rather than a pure print-first workflow. [S20] [S21]
• Choose 3D Scanner App if you want a simple LiDAR utility with common export formats and will do cleanup elsewhere. [S24]
• Choose Heges if you specifically want to compare close-range TrueDepth capture against rear LiDAR behavior on iPhone. [S25]

Best 3D scanner app for Android in 2026

Android’s biggest strength is breadth. Google says over 88% of active devices supported ARCore Depth as of May 2026, which means many Android phones can participate in depth-aware scanning or AR scene understanding even without a dedicated LiDAR scanner. [S09] ARCore Depth is often computed via depth-from-motion from the RGB camera, with optional ToF fusion, so it isn’t equivalent to Apple’s dedicated rear LiDAR scanner. [S10] That is the key Android caveat: Android is often excellent for photogrammetry and increasingly capable for scene capture, but the hardware and reconstruction path vary more from device to device.

For most Android users, the practical shortlist starts with KIRI Engine, RealityScan Mobile, and Scaniverse. KIRI is the most format-flexible official option here for photo-based workflows. [S14] RealityScan is strong for textured object capture on Android 7/API 24+ devices with ARCore support, with GLB local model storage and Sketchfab downloads in FBX, USDZ, and GLB or glTF. [S17] [S18] [S19] Scaniverse is attractive when you want mesh-or-splat scene outputs and are comfortable checking mode-specific behavior at publication time. [S20] [S21] Polycam belongs on Android only within its verified scope, because its own support page says Android Space Mode is not available. [S13]

Best photogrammetry app for 3D scanning

Photogrammetry usually beats depth-first capture when the subject is a small, textured object and the goal is a detailed textured asset. That is especially true when the surface gives the solver plenty of visual features to match from image to image. In those cases, you care less about room-scale depth sensing and more about overlap, consistent lighting, and a clean textured-mesh output for Blender, a game engine, or post-processing.

Among the apps in this list, RealityScan Mobile is the clearest choice when you want a straightforward textured-asset pipeline with Sketchfab distribution and common download formats afterward. [S18] [S19] KIRI Engine is the most flexible when you need direct format options such as OBJ, FBX, STL, GLB, GLTF, USDZ, PLY, or XYZ from photo-based workflows, and it is also the main app here that explicitly documents a Featureless Object Scan mode. [S14] Polycam can also work well for object photogrammetry, but its object workflow is cloud-processed and its export access changes by tier. [S11] [S12] Scaniverse belongs in this discussion mostly when your end goal is a visual mesh-or-splat result rather than a conventional textured-asset mesh. [S20] [S21] [S22]

Table note: This comparison is about photogrammetry deliverables, not room LiDAR capture. [S11] [S12] [S14] [S18] [S19] [S20] [S21]

When a phone photo set is not enough, the same job can move to the desktop. AI Photogrammetry in Artec Studio turns photos or video from any camera into a textured mesh with real geometry underneath, rather than a surface that only looks right from a distance, which is the difference that matters when the model has to be measured, reverse-engineered, or printed. [S31] For a photo-and-video-only workflow without a scanner, Artec Studio Lite is the lighter desktop option; it is desktop software, not a mobile app. [S32]

Workflow: from scan to mesh, OBJ, STL, or GLB

A useful mobile scan is not a one-button event. It is a chain: capture, processing, cleanup, scaling, export, and validation.

1. Prep object and lighting.
2. Capture with overlap or stable motion.
3. Process the scan, noting whether that mode is on-device, cloud, or hybrid.
4. Clean the mesh, separately from checking dimensions.
5. Establish scale with a known reference.
6. Export: OBJ, GLB, or USDZ for textured assets; STL for geometry-only print workflows.
7. Validate and repair before printing or measurement use.

The step many beginners skip is scale establishment and validation. Cleanup makes a model easier to use, but it does not prove the dimensions are trustworthy. Independent research on photogrammetry has shown that adding more and better-placed scale bars can materially improve results: one construction study reported its four-scale-bar trial produced an average absolute difference of 1.72 mm, while a later ISPRS paper reported a 67% accuracy improvement from calibrated scale-bars and, in that study context without RTK, reduced local distance error from 1–3 m to less than 4 cm. [S29] [S30] The exact numbers belong to those study setups, not to every phone scan, but the workflow lesson is broadly useful: if dimensions matter, add a known reference and then check the result.

Choose export formats by downstream use, not by habit. OBJ, GLB, and USDZ are common handoff formats when you need textured assets or AR viewing. STL does not carry textures. Point-cloud-style outputs are different again: ASTM’s E57 file-format specification notes that an E57 file can store 3D point data, attributes such as color or intensity, and 2D imagery, which is why a point-cloud workflow should not be collapsed into an STL-style mesh workflow. [S04] If your model is headed to a slicer, think geometry and repair. If it is headed to Blender or a game engine, think textures, normals, and scene compatibility.

Performance metrics that matter

The most abused word in mobile scanning is still accuracy. For practical use, separate at least six ideas: truth accuracy against a reference, precision or repeatability across repeated captures, resolution or point density, tracking robustness, texture quality, and processing location or privacy behavior. Surface physics matter too. NIST’s range-performance work lists range, angle-of-incidence, reflectivity, azimuth angle, method of obtaining the range measurement, and target type as factors that affect performance. [S06] That is why black, shiny, angled, or low-feature surfaces can behave badly even when the app is otherwise competent.

Don’t compare using only:

• capture method plus hardware dependency
• processing location, which can be mode-dependent
• export formats, which can be mode- or tier-dependent
• scale control and validation steps
• repeatability versus truth accuracy
• downstream cleanup needs

Independent research is useful here mainly as a warning against overgeneralization. A 2023 study comparing iPhone 13 Pro room-scanning apps found that mean deviations and RMSE varied widely across apps within that specific setup. [S26] A 2024 PLOS ONE study, by contrast, reported very high agreement metrics for smartphone scanning in a controlled residual-limb clinical setup, but that result is tightly scoped to the tested workflow and subject class. [S27] A 2024 ISPRS study also benchmarked multiple Samsung Galaxy S-series phones and iPhones against an ATOS 5 structured-light reference in controlled conditions, reinforcing the point that phones can perform well, but only in methods-and-scene contexts that must be stated clearly. [S28]

Standards, verification, and “accuracy” language

ASTM terminology matters because it keeps people from comparing unlike things with the same words. ASTM E2544-24 is the active Standard Terminology for Three-Dimensional (3D) Imaging Systems, and its store page lists 13 pages and DOI 10.1520/E2544-24. [S01] That is a terminology standard, not a certificate that any app is accurate. The wider standards context goes back to ASTM’s announcement of Committee E57 on July 1, 2006, and NIST’s later summary that E57 was established in 2006 with four subcommittees: Terminology, Test Methods, Best Practices, and Data Interoperability. [S02] [S03]

For verification, the more relevant model is a framework such as VDI/VDE 2634 Part 2. The DIN Media listing describes it as defining acceptance and monitoring procedures and test bodies for imaging optical 3D measuring systems using area scanning, limited here to single-view triangulation systems, and lists issue date 2012-08 with 16 pages. [S05] Consumer phone apps rarely publish this kind of acceptance-test evidence for their end-to-end workflows. So when app listings talk about precision, the technically safer reading is usually that they are describing a mode or internal setting, not presenting a standards-based validation result.

Limitations & failure cases

Mobile scanning fails for physical reasons before it fails for branding reasons. Reflective and transparent materials confuse depth and image-based reconstruction in different ways. Dark or absorptive surfaces can weaken returned signal. Thin geometry such as chair legs can vanish or wobble. Featureless walls and repetitive patterns can undermine photogrammetric matching. Moving subjects break overlap assumptions. Fast camera motion causes tracking loss. NIST’s range-performance work is a good reminder that range, angle-of-incidence, reflectivity, azimuth angle, measurement method, and target type all change the result. [S06]

When the deliverable must support formal inspection, controlled dimensional acceptance, or regulated documentation, that is usually where dedicated structured-light or metrology hardware becomes the better fit. That is not because phone apps are useless, but because standards-style acceptance frameworks exist for optical measuring systems in a way consumer app workflows typically do not document publicly. [S05]

Applications: 3D printing, AR, games, design, heritage, and rooms

For 3D printing, phone scans are usually best treated as reference meshes or starting geometry, not as finished production parts. An STL export may be convenient, but it does not guarantee watertightness, correct wall thickness, clean manifold geometry, or trustworthy scale. In many maker workflows, the better move is to scan for reference, remodel in CAD or Blender, then print the cleaned result.

For games and AR, textured mesh formats matter more. RealityScan’s documented Sketchfab pipeline and KIRI’s broad photo-based exports make them strong choices when you want assets that move into engines or viewers with minimal format friction. [S14] [S19] For visual capture, splats are increasingly compelling because they preserve scene appearance efficiently. Niantic’s SPZ format is explicitly a compressed 3D Gaussian splat format, and the underlying 3D Gaussian Splatting literature frames splats as a radiance-field rendering representation rather than a normal editable polygon mesh. [S22] [S23]

For rooms, design documentation, heritage recording, and AEC-style walkthroughs, the advantage of phone apps is speed and accessibility. Scaniverse’s Sites workflow and LiDAR-first iPhone apps make fast capture plausible for many non-regulated tasks. [S20] [S24] But fast documentation is not the same as a survey deliverable or an acceptance-tested measurement record.

Conclusion: choosing among the best 3D scanner apps in 2026

Choosing among the best 3D scanner apps in 2026 comes down to identifying your job, hardware, and deliverable class before you install anything. If you have a LiDAR iPhone and care about rooms, start with Polycam or 3D Scanner App. If you care about textured objects, start with RealityScan Mobile or KIRI Engine. If you want visual scene capture and splats, Scaniverse and KIRI deserve more attention than print-oriented lists usually give them. If you are comparing TrueDepth and LiDAR behavior on iPhone, Heges is a useful example app, but not a substitute for independent validation.

The last check should always be dimensional trust. A clean mesh is not the same as a verified one, and an STL export is not the same as a print-ready part. If tolerance matters, add a known reference, validate the result, and re-check current app capabilities before committing to a workflow.

If you have already hit the limits of phone capture, Artec’s own guide to 3D scanner apps covers the same ground from the hardware side, including where a handheld scanner takes over from an app. [S33]

FAQ

What is the best 3D scanner app in 2026?

There is no single best app for every task. Polycam is strongest for LiDAR room capture on iPhone, RealityScan Mobile is strongest for many textured-object workflows, KIRI Engine is the most format-flexible cross-method option, and Scaniverse is especially relevant for visual mesh-or-splat scene capture. [S11] [S14] [S19] [S20]

What is the best 3D scanner app for iPhone in 2026?

For LiDAR-equipped iPhones, Polycam is the safest starting point for room capture because its LiDAR-oriented modes process locally on-device. [S11] [S13] For non-LiDAR object capture, RealityScan Mobile and KIRI Engine are usually better first installs. [S14] [S17]

What is the best 3D scanner app for Android in 2026?

For most Android users, start with KIRI Engine for flexible exports, RealityScan Mobile for textured objects, and Scaniverse for scene capture and splats. [S14] [S17] [S21] Polycam on Android should be treated as more limited, because its own support page says Android Space Mode is not available. [S13]

Do Android phones have LiDAR for 3D scanning?

Usually not, in the Apple sense. Many Android devices support ARCore Depth instead, and Google says ARCore Depth often relies on depth-from-motion from the RGB camera with optional ToF fusion rather than a dedicated rear LiDAR scanner. [S09] [S10]

What’s the best photogrammetry app for 3D scanning?

RealityScan Mobile is often the easiest pick for textured objects and Sketchfab-centered sharing, while KIRI Engine is the better pick when you need more direct control over export formats or want its documented Featureless Object Scan path. [S14] [S18] [S19]

Expert: How do I validate scale or measurement accuracy from a mobile scan?

Use a known reference during capture, then measure the resulting model against that reference after processing. Independent studies on photogrammetry showed better results when scale bars were added, including one construction experiment where the four-scale-bar trial reported an average absolute difference of 1.72 mm, and an ISPRS study that reported a 67% accuracy improvement from calibrated scale-bars in its workflow. [S29] [S30]

Expert: What standards exist for verifying optical 3D measuring systems, and do phone apps comply?

ASTM E2544-24 is an active terminology standard for 3D imaging systems, which helps with consistent language but is not an app certification. [S01] VDI/VDE 2634 Part 2 is closer to a verification framework because it defines acceptance and monitoring procedures and test bodies for certain imaging optical 3D measuring systems. [S05] Consumer phone apps generally do not publish end-to-end acceptance-test results of that kind.

Sources

• S01: ASTM E2544-24 Standard Terminology for Three-Dimensional (3D) Imaging Systems — https://store.astm.org/e2544-24.html
• S02: ASTM press release: 3D Imaging Technology Is Focus of New ASTM Standards Initiative — https://www.astm.org/news/press-releases/3d-imaging-technology-focus-new-astm-standards-initiative
• S03: NIST: ASTM E57 3D Imaging Systems Committee: An Update — https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=860707
• S04: ASTM E2807-11R19 Standard Specification for 3D Imaging Data Exchange, Version 1.0 — https://store.astm.org/e2807-11r19.html
• S05: DIN Media listing for VDI/VDE 2634 Part 2 — https://www.dinmedia.de/de/technische-regel/vdi-vde-2634-blatt-2/154896837
• S06: NIST TN 1695: Characterization of the Range Performance of a 3D Imaging System — https://www.nist.gov/publications/characterization-range-performance-3d-imaging-system-nist-tn-1695
• S07: Apple ARKit supportsSceneReconstruction(_:) documentation — https://developer.apple.com/documentation/ARKit/ARWorldTrackingConfiguration/supportsSceneReconstruction%28_%3A%29?changes=_8%2C_8
• S08: Apple ARKit ARDepthData documentation — https://developer.apple.com/documentation/arkit/ardepthdata
• S09: Google ARCore supported devices page — https://developers.google.cn/ar/devices?hl=en
• S10: Google ARCore Depth overview — https://developers.google.com/ar/develop/depth
• S11: Polycam help: Do I Need Wi‑Fi/LTE to Process My Captures? — https://learn.poly.cam/hc/en-us/articles/30549172696084-Do-I-Need-Wi-Fi-LTE-to-Process-My-Captures
• S12: Polycam help: What File Types Can Polycam Export? — https://learn.poly.cam/hc/en-us/articles/27756102599572-What-File-Types-Can-Polycam-Export
• S13: Polycam help: Which Devices Are Supported by Polycam? — https://learn.poly.cam/hc/en-us/articles/34419168797972-Which-Devices-Are-Supported-by-Polycam
• S14: KIRI Engine export formats — https://www.kiriengine.app/features/export-formats
• S15: KIRI Engine blog: How to Capture 3D Gaussian Splats with KIRI Engine — https://www.kiriengine.app/blog/how-to-capture-3d-gaussian-splats-kiri-engine
• S16: KIRI Engine blog: What Is AI-Enhanced LiDAR? — https://www.kiriengine.app/blog/what-is-ai-enhanced-lidar
• S17: RealityScan Mobile official page — https://www.realityscan.com/mobile
• S18: Epic documentation: RealityScan Project Files — https://dev.epicgames.com/documentation/realityscan-mobile/RealityScan-Project-Files?lang=en-US
• S19: Unreal Engine RealityScan page — https://www.unrealengine.com/en-US/realityscan
• S20: Niantic Spatial Scaniverse Quickstart for Sites — https://www.nianticspatial.com/docs/scaniverse/quickstart/
• S21: Scaniverse Google Play listing — https://play.google.com/store/apps/details?id=com.nianticlabs.scaniverse&hl=en
• S22: Niantic SPZ GitHub repository — https://github.com/nianticlabs/spz
• S23: Kerbl et al., 3D Gaussian Splatting for Real-Time Radiance Field Rendering — https://arxiv.org/abs/2308.04079
• S24: 3D Scanner App official site — https://www.3dscannerlidar.com/
• S25: Heges 3D Scanner App Store listing — https://apps.apple.com/us/app/heges-3d-scanner/id1382310112
• S26: Geomatics 2023: Use of Smartphone Lidar Technology for Low-Cost 3D Building Documentation with iPhone 13 Pro — https://www.mdpi.com/2673-7418/3/4/30/html
• S27: PLOS ONE 2024: Smartphone scanning is a reliable and accurate alternative to contemporary residual limb measurement techniques — https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0313542
• S28: ISPRS Archives 2024 smartphone photogrammetry comparison study — https://isprs-archives.copernicus.org/articles/XLVIII-2-W8-2024/211/2024/
• S29: Dai, Feng, and Hough 2014: Photogrammetric error sources and impacts on modeling and surveying in construction engineering applications — https://link.springer.com/article/10.1186/2213-7459-2-2
• S30: ISPRS 2021: Improving the Internal Accuracy of UAV-Image Blocks Using Local Low-Altitude Flights and Scale-Bars — https://isprs-archives.copernicus.org/articles/XLIII-B1-2021/183/2021/isprs-archives-XLIII-B1-2021-183-2021.pdf
• S31: Artec 3D: AI Photogrammetry in Artec Studio — https://www.artec3d.com/3d-software/artec-studio/photogrammetry
• S32: Artec 3D: Artec Studio Lite — https://www.artec3d.com/3d-software/artec-studio/lite
• S33: Artec 3D Learning Center: Best 3D scanner apps — https://www.artec3d.com/learning-center/best-3d-scanner-apps