This is Changing 3D Scanning!!
TLDRThe video discusses the advancements in 3D scanning and augmented reality, highlighting the Kiri engine as a notable tool for both iPhone and Android users. The Kiri engine, which was previously reviewed as an alternative to polycam, now features a new virtual scanning technique using NeRF (Neural Radiance Fields) technology, specifically the norlandulo method. This technology overcomes some limitations of photogrammetry, such as difficulties scanning reflective surfaces or objects with a single color. The video also touches on the potential of NeRF and norlandulo to revolutionize 3D scanning, despite not being a direct replacement for photogrammetry. The Kiri engine's featureless object mode, based on Orlando, is particularly powerful and offers a new way to create 3D models from videos. The video concludes by inviting viewers to try the Kiri engine for free and subscribe for more informative content.
Takeaways
- ๐ 3D scanning and photogrammetry are gaining more attention with the rise of AI and AR technologies.
- ๐ฑ Kiri Engine is a versatile tool for 3D scanning and AR technology, available for both iPhone and Android.
- ๐ Kiri Engine has introduced a new feature called virtual scanning using NeRF (Neural Radiance Fields) technology.
- ๐ NeRF technology is capable of reconstructing 3D scenes from a set of images or videos, even those without distinct features.
- ๐ Norlandulo is an emerging technique that improves upon NeRF by generating high-quality 3D models usable in software like Blender.
- ๐ซ Photogrammetry has limitations with objects that lack distinct features, such as single-color or reflective surfaces.
- ๐ Kiri Engine's Norlandulo-based featureless object mode is powerful and addresses the limitations of photogrammetry.
- ๐ฅ To use the featureless scanning mode, you record a video of the object, moving around it to capture it from different angles.
- โณ The video length for scanning is limited to 2 minutes due to server limitations, but this duration is typically sufficient.
- ๐ The processing time for featureless scanning may be longer than photogrammetry as it is a newer technology with ongoing optimization.
- ๐ Kiri Engine offers more than just featureless scanning, including a web app, access library, in-app editing, and texture toning.
Q & A
What is the current status of 3D scanning in the industry?
-3D scanning is one of the most challenging subjects in the 3D field right now, gaining more attention every day due to the market's shift towards artificial intelligence and augmented reality.
What is the Kiri engine and how does it relate to 3D scanning?
-The Kiri engine is a tool for 3D scanning and AR technology that was reviewed as a great alternative to iPhone 3D scanning apps like Polycam. It is now available for Android as well and supports photogrammetry.
What is the new feature in Kiri engine that the speaker is particularly interested in?
-The speaker is particularly interested in the new feature of virtual scanning using NeRF technology, specifically a technique called Norlandulo, which aims to mitigate some of the shortcomings of photogrammetry.
How does NeRF technology work and what is its advantage over photogrammetry?
-NeRF, which stands for neural Radiance Field, is a machine learning technology that uses AI to reconstruct an entire 3D scene from a set of images or a video. It can reconstruct scenes and objects that lack enough feature points for photogrammetry, making it more versatile.
What is Norlandulo and how does it improve upon NeRF technology?
-Norlandulo is an advancement over NeRF technology that not only generates neural rendered scenes but also produces a high-quality mesh of the object that is usable in 3D software like Blender.
What are the limitations of photogrammetry when it comes to scanning certain objects?
-Photogrammetry struggles with objects that don't have many feature points, such as objects with a single color, transparent or translucent materials, or extremely reflective surfaces, because there aren't enough details for the algorithm to create a complete shape.
How does the Kiri engine's featureless object mode work?
-To use the featureless object mode in Kiri engine, you click the big plus button, select the featureless option, and instead of taking photos, you record a video of the object. You then circle around the object, trying to keep the recording steady and consistent, and submit the video to the cloud for processing.
What is the processing time for scanning with the Kiri engine's featureless mode compared to photogrammetry?
-The processing time for the featureless mode might be slightly longer than photogrammetry due to it being a newer feature. However, more optimization is expected to be on the way.
Will NeRF technology eventually replace photogrammetry?
-No, NeRF is not expected to replace photogrammetry. While it can handle objects that photogrammetry cannot, it currently can't produce high-quality 3D models and textures suitable for a typical 3D pipeline.
What are the other features of the Kiri engine apart from the featureless mode?
-Apart from the featureless mode, Kiri engine offers a web app, access to a library, the ability to create a floor plane, in-app editing, texture toning, and more.
How can one try the Kiri engine for free?
-You can try the Kiri engine for free by downloading it from the link provided in the video description. It is compatible with both Android and iPhone devices.
Outlines
๐ Introduction to 3D Scanning and Kiri Engine
The first paragraph introduces 3D scanning as a challenging and growing field within the 3D industry, gaining attention due to advancements in artificial intelligence (AI) and augmented reality (AR). It highlights the Kiri Engine as a versatile tool for 3D scanning and AR technology, compatible with both iPhone and Android devices, and capable of integrating with photogrammetry. The paragraph also discusses the Kiri Engine's new feature for virtual scanning using NeRF (Neural Radiance Fields) technology, which is designed to overcome some limitations of photogrammetry, such as difficulties scanning reflective or black surfaces. The Kiri Engine is presented as a solution for creating 3D models from images for various applications like gaming, movies, and medical fields.
๐ NeRF and Norlandulo: New Horizons in 3D Scanning
The second paragraph delves into the limitations of photogrammetry and introduces NeRF (Neural Radiance Fields) as a machine learning technology that uses AI to reconstruct 3D scenes from a series of images or videos. It emphasizes NeRF's ability to handle scenes and objects with insufficient feature points for traditional photogrammetry. The paragraph then introduces Norlandulo, a more advanced technique that not only generates neural-rendered scenes like NeRF but also produces 3D models with high mesh quality usable in software like Blender. The Kiri Engine's implementation of Norlandulo is discussed, with instructions on how to use the featureless object mode for scanning objects by recording a video and submitting it to the cloud. The paragraph concludes by addressing the question of whether NeRF will replace photogrammetry, stating that while NeRF is powerful, it currently cannot handle high-quality 3D models and textures required for typical 3D pipelines, but Norlandulo 2.0 shows potential in this regard.
Mindmap
Keywords
๐ก3D Scanning
๐กPhotogrammetry
๐กKiri Engine
๐กNeRF (Neural Radiance Fields)
๐กNerf Technology
๐กOrlando
๐กFeatureless Object Mode
๐กPoint Cloud
๐กAR (Augmented Reality)
๐กAI (Artificial Intelligence)
๐กVFX (Visual Effects)
Highlights
3D scanning is becoming increasingly important in the industry with growing attention on technologies like photogrammetry.
The market is shifting towards artificial intelligence and augmented reality, making 3D scanning tools more accessible.
Kiri Engine is a notable tool for 3D scanning and AR technology, offering an alternative to iPhone 3D scanning apps like Polycam.
Kiri Engine now supports Android and has QR code support for those without the hardware for Real Time 3D scanning.
A new feature in Kiri Engine is virtual scanning using NeRF technology, specifically a technique called Norlandulo.
NeRF technology aims to address some of the limitations of photogrammetry, such as difficulties scanning reflective surfaces or black colors.
Kiri Engine's future scan and Norlandulo aim to improve mesh quality, which is a current shortcoming of NeRF.
Photogrammetry requires distinct features in an image to create a complete shape, which can be a limitation.
NeRF technology does not rely on distinct features and can generate decent meshes for surfaces rich in texture.
Objects with few feature points, such as single-colored or transparent objects, often fail with photogrammetry.
Kiri Engine is available for free and compatible with both Android and iPhones.
NeRF stands for Neural Radiance Field, a machine learning technology that uses AI to reconstruct 3D scenes from images or videos.
Norlandulo is an advancement over NeRF, generating neural rendered scenes and usable 3D objects.
Norlandulo is based on recent research by Nvidia and is being implemented in Kiri Engine.
To use the featureless object mode in Kiri Engine, record a video of the object and submit it to the cloud for processing.
Kiri Engine limits video length to 2 minutes for featureless scanning due to server limitations.
NeRF is not expected to replace photogrammetry, but Norlandulo 2.0 may be a stronger contender as it addresses the shortcomings of both.
Kiri Engine offers more than just featureless mode, including a web app, access library, in-app editing, and more.