Comparing Top Five 3D Scanner Apps | Photogrammetry VS NeRFs VS LiDAR
TLDRIn this video, Jack reviews the top five 3D scanning apps from an industry perspective, addressing the question of differences between apps like K engine and others. The video discusses three main technologies used in 3D scanning: photogrammetry, NeRF, and LiDAR, and how they are employed by the five apps reviewed. The apps include K engine, Polycam, Reality Scan, Luma, and Scanny Verse, each with their own advantages and limitations. Jack tests the apps on various objects and compares their capabilities, such as capturing details, exporting materials, and handling different surfaces. He concludes that K engine stands out for its combination of photogrammetry and NeRF, offering versatility and quality, while also noting the strengths of other apps like Polycam and Scanny Verse in specific scenarios. The summary provides viewers with insights to help them choose the right app for their 3D scanning needs.
Takeaways
- 📱 **Five 3D Scanner Apps Compared**: Jack reviews the top 3D scanning apps, discussing their differences and suitability for various projects.
- 🔍 **Three Core Technologies**: Photogrammetry, NeRF, and LiDAR are the main technologies used for 3D scanning on smartphones.
- 📐 **Photogrammetry Details**: This method uses feature points on an object's surface to create 3D point clouds and models, suitable for objects with enough detail.
- 📈 **NeRF's Advantages**: NeRF, utilizing neural networks, can scan objects with insufficient feature points, including shiny surfaces.
- 🚀 **LiDAR's Real-Time Feedback**: LiDAR technology provides instant 3D scanning results on supported iPhone Pro models without the need for cloud processing.
- 🔧 **KY Engine's Versatility**: KY Engine uses both photogrammetry and NeRF, making it capable of scanning complex and simple objects without hardware limitations.
- 📉 **Polycam and Scannyverse with LiDAR**: These apps offer LiDAR scanning capabilities but may struggle with simple or shiny objects.
- 🔄 **Object Masking and Bottom Capturing**: KY Engine and Polycam can capture the bottom of 3D models using object masking, a feature not available in Reality Scan.
- 🎨 **PBR Materials and Export Options**: KY Engine can export PBR materials and offers an additional map export, while Reality Scan cannot capture the model's bottom without object masking.
- 🤖 **Neural Surface Reconstruction**: KY Engine employs neural surface reconstruction, enhancing the quality of models produced through NeRF technology.
- 📲 **Device Limitations**: Scanning apps like Luma are restricted to iOS devices and may not capture the bottom of 3D models, whereas KY Engine is available on both iOS and Android.
Q & A
What are the three major technologies used for 3D scanning on smartphones?
-The three major technologies used for 3D scanning on smartphones are photogrammetry, NeRF, and LiDAR.
How does photogrammetry work in the context of 3D scanning?
-Photogrammetry uses feature points on an object's surface as a reference to construct 3D point clouds and, in turn, to construct a 3D model of an object.
Which apps use photogrammetry technology?
-The apps that use photogrammetry technology are Cur. Engine, Polycam, and Reality Scan.
What is the main advantage of NeRF in 3D scanning?
-The main advantage of NeRF is that it introduces neural networks into the scanning process, allowing it to scan objects that do not have enough feature points, even shiny objects.
Which apps were tested using NeRF technology?
-The apps tested using NeRF technology were Cur. Engine and Luma.
What is LiDAR technology and which devices support it?
-LiDAR technology uses invisible laser beams to scan the surrounding environment and create 3D models by measuring the reflected light. It is only available on iPhone Pro models that have the LiDAR hardware.
Which apps utilize LiDAR capabilities?
-Polycam and Scanny Verse are the apps that utilize LiDAR capabilities.
What are some limitations of LiDAR when scanning objects?
-LiDAR won't perform well if the object has a simple structure or a shiny and reflective surface, and it has fewer face counts compared to photogrammetry.
Which app can export 3D models as retopologized quad mesh?
-Cur. Engine is the only app that can export 3D models as retopologized quad mesh.
What is the main difference between Cur. Engine and Luma when it comes to NeRF technology?
-Cur. Engine uses a technology called neural surface reconstruction, which is optimized for high-quality models, while Luma uses traditional NeRF technology.
What is the main advantage of Scanny Verse over other apps when using LiDAR?
-Scanny Verse uses LiDAR on iPhone Pro models and structure from motion for other iPhones, allowing it to achieve less impressive but still good results without LiDAR hardware.
What is the recommendation for someone who cares about the details of the 3D model when using photogrammetry?
-For someone who cares about the details of the 3D model when using photogrammetry, Cur. Engine and Reality Scan are recommended.
Outlines
📱 3D Scanning App Review Overview
Jack introduces his video with the intent to review and compare his top five 3D scanning apps. He addresses the frequent inquiries about the differences between various 3D scanning apps, particularly the K engine. The video aims to clarify the distinctions between these apps to help viewers select the most suitable one for their projects. Jack begins by discussing the three primary technologies used in 3D scanning on smartphones: photogrammetry, NeRF, and LiDAR. He notes that the five apps under review utilize these technologies differently, leading to varying advantages and limitations. The video includes a hands-on test of photogrammetry using three apps on an object with ample feature points. Despite initial appearances, the resulting 3D models from each app show significant differences, as illustrated in a comparative chart. Jack highlights the capabilities of each app, such as object masking and PBR material export, and recommends K engine and Reality Scan for their superior performance in generating detailed 3D models.
🚀 NeRF Technology and its Applications
The script shifts the focus to NeRF, a newer scanning technique that employs neural networks to enhance the scanning process, especially for objects lacking distinct features. Unlike photogrammetry, NeRF can scan even shiny objects. However, it has its drawbacks, such as the inability to use object masking and the requirement for more computational power, leading to longer processing times and higher costs. Jack presents a comparison between the models generated by two apps using NeRF technology, noting the stark differences due to the additional neural surface reconstruction employed by K engine. He also touches upon the challenges of scanning a shiny object with photogrammetry, which proves to be a nightmare scenario. The paragraph concludes with a teaser for the next technology to be tested.
Mindmap
Keywords
💡3D Scanner Apps
💡Photogrammetry
💡NeRF (Neural Radiance Fields)
💡LiDAR
💡Object Masking
💡PBR Materials
💡Quad Mesh
💡Face Counts
💡Neural Surface Reconstruction
💡Structure from Motion
💡Hardware Limitations
Highlights
Jack reviews the top five 3D scanning apps from an industry perspective.
Three major technologies for 3D scanning on smartphones: photogrammetry, NeRF, and LiDAR.
KY Engine uses photogrammetry and NeRF, while other apps use different combinations of these technologies.
Photogrammetry relies on feature points to construct 3D point clouds and models.
KY Engine and Polycam can capture the bottom of 3D models using object masking features, unlike Reality Scan.
KY Engine is the only app that can export 3D models as retopologized quad mesh.
KY Engine and Reality Scan can generate more face counts than Polycam for detailed 3D models.
NeRF introduces neural networks to 3D scanning, capable of scanning objects with insufficient feature points.
NeRF has limitations, such as the inability to use object masking and longer processing times.
KY Engine uses neural surface reconstruction for optimized and high-quality NeRF models.
LiDAR technology uses laser beams for real-time 3D scanning feedback.
Polycam and Scanny Verse offer LiDAR capabilities, but are limited to iPhone Pro models.
LiDAR performs well for objects with many features but struggles with simple, shiny surfaces.
KY Engine is versatile, combining photogrammetry and NeRF without hardware limitations.
Scanning Verse is user-friendly, processing scans locally on the iPhone but with fewer face counts.
Reality Scan has a fancy UI and uploads photos during scanning to save time.
Luma excels at scanning both complex and simple objects using NeRF technology but lacks the ability to capture the bottom of 3D models.
KY Engine is recommended for superior mesh quality and availability on both iOS and Android.