This is Changing 3D Scanning!!

InspirationTuts
25 Aug 202309:40

TLDRThe script discusses the advancements in 3D scanning and augmented reality, highlighting Kiri Engine as a versatile tool for both iPhone and Android. It introduces Neurological technology, specifically the Norlandulo technique, which overcomes photogrammetry's limitations by using AI to reconstruct 3D models from images or video. Kiri Engine's featureless scanning leverages this technology, though it's noted that while Nerf may not replace photogrammetry entirely, the emerging Norlandulo method shows promise in addressing its shortcomings.

Takeaways

  • 🌟 3D scanning and photogrammetry are gaining increased attention in the industry, especially with the rise of AI and AR technologies.
  • 📱 Kiri Engine is a versatile 3D scanning and AR tool, offering an alternative to popular iPhone 3D scanning apps like Polycam and is also available for Android devices.
  • 🚀 Kiri Engine recently introduced a new feature called virtual scanning, which utilizes Nerf technology to overcome some limitations of traditional photogrammetry.
  • 🎥 Nerf stands for Neural Radiance Field, a machine learning technique that reconstructs 3D scenes from a set of images or a video, making it possible to scan objects with insufficient feature points for photogrammetry.
  • 🤖 The advancement of technology has led to the development of neurological, a technique that not only generates neural rendered scenes like Nerf but also produces usable 3D models for software like Blender.
  • 🔍 Kiri Engine's featureless scanning option allows users to scan objects without relying on distinct features, addressing the limitations of photogrammetry with objects like reflective or single-colored surfaces.
  • 📹 To use the featureless scanning feature in Kiri Engine, users need to record a video of the object from various angles, with a maximum length of 2 minutes due to server limitations.
  • ⏳ Processing time for the featureless scan may be longer than traditional photogrammetry, as it is a newer technology that requires optimization.
  • 🔥 The combination of Nerf and neurological technologies in Kiri Engine makes it a powerful tool for 3D scanning, potentially addressing most of the shortcomings of both techniques.
  • 🛠️ Apart from the featureless mode, Kiri Engine offers additional functionalities like an access library, in-app editing features, and the ability to create a floor plane, enhancing the overall 3D scanning experience.

Q & A

  • What is 3D scanning and how does it relate to photogrammetry?

    -3D scanning is the process of converting 2D images into a three-dimensional model, and photogrammetry is a method used in 3D scanning that involves taking multiple photos of an object or environment to create a detailed 3D representation. It is widely used in various fields such as entertainment, medical, and engineering.

  • What is Kiri Engine and how does it differ from other 3D scanning apps?

    -Kiri Engine is a versatile 3D scanning and augmented reality tool that offers multiple ways to create 3D models, including traditional photogrammetry and the newly emerging NeRF technology. It differs from other apps as it is available for both iPhone and Android, and it is one of the few platforms that currently offer the NeRF technology for 3D scanning.

  • What are the limitations of photogrammetry in 3D scanning?

    -Photogrammetry requires distinct features in the images, such as changes in color or shape, to create a complete 3D model. It may struggle with objects that have limited feature points, such as single-colored objects, transparent or translucent materials, and highly reflective surfaces.

  • What is NeRF technology and how does it improve upon photogrammetry?

    -NeRF, or neural Radiance Field, is a machine learning technology that uses artificial intelligence to reconstruct 3D scenes from a set of images or a video. It can create 3D models of scenes and objects that lack enough feature points for traditional photogrammetry, making it capable of handling more challenging materials and surfaces.

  • What is the Norlandulo technique and how does it compare to NeRF?

    -Norlandulo is an advanced technique based on NeRF that not only generates neural rendered scenes but also produces high-quality meshes that are usable in 3D software like Blender. It addresses some of the shortcomings of NeRF, particularly in terms of mesh quality and usability in standard 3D pipelines.

  • How does Kiri Engine implement Norlandulo technology?

    -Kiri Engine has integrated the Norlandulo technology into its platform, making it the only app currently available on both iPhone and Android that allows users to try Norlandulo for 3D scanning. This feature is accessed through a 'featureless' option in the app, where users can record a video of the object they want to scan.

  • What are the advantages of using Kiri Engine's featureless scanning mode?

    -The featureless scanning mode in Kiri Engine, powered by Norlandulo, allows users to scan objects that might be difficult for traditional photogrammetry, such as objects with reflective surfaces or lack of distinct features. It also has the potential to improve mesh quality and make the 3D models more suitable for various applications.

  • How long should the video be for using Kiri Engine's featureless scanning mode?

    -Kiri Engine limits the video length to 2 minutes for featureless scanning due to server limitation purposes. This duration should be sufficient for most scanning tasks.

  • What is the processing time for Kiri Engine's featureless scanning mode?

    -The processing time for the featureless scanning mode in Kiri Engine might be slightly longer than traditional photogrammetry due to the complexity of the NeRF and Norlandulo technologies involved. However, optimizations are ongoing to improve this aspect.

  • Will NeRF and Norlandulo replace photogrammetry in the future?

    -While NeRF shows promise in handling challenging materials and surfaces, it currently cannot replace photogrammetry for creating high-quality 3D models and textures. However, Norlandulo, with its improved mesh quality and usability, could potentially become a stronger contender for replacing photogrammetry in certain applications.

  • What other features does Kiri Engine offer besides the featureless mode?

    -Besides the featureless mode, Kiri Engine offers a web app, an access library, the ability to create a floor plane, in-app editing, texturing, and many other features to enhance the 3D scanning and modeling experience.

Outlines

00:00

📱 Advancements in 3D Scanning and AR Technologies

This paragraph discusses the growing importance of 3D scanning and photogrammetry in the industry, with a focus on the increasing interest in artificial intelligence and augmented reality. It introduces Kiri Engine, a tool for 3D scanning and AR technology, as a versatile alternative to other iPhone 3D scanning apps. The paragraph highlights the new feature of virtual scanning using NeRF technology, which was recently covered in a video. It contrasts NeRF's ability to mitigate some of the shortcomings of photogrammetry, such as difficulties with reflective surfaces or black objects, and discusses the potential improvements with Kiri Engine's future updates.

05:01

🤖 NeRF and Norlandulo: The Future of 3D Modeling

The second paragraph delves into the details of NeRF (Neural Radiance Field) technology and its role in 3D modeling. It explains that NeRF is a machine learning technique that reconstructs 3D scenes from a set of images or video, making it particularly effective for objects with insufficient feature points for photogrammetry. The paragraph introduces Norlandulo, an advancement over NeRF, which not only generates neural rendered scenes but also creates usable 3D models in software like Blender. It describes how Kiri Engine has implemented Norlandulo technology, making it accessible on both iPhone and Android devices. The paragraph concludes by discussing the limitations of photogrammetry and how NeRF and Norlandulo technologies aim to address these, suggesting that while NeRF may not fully replace photogrammetry, Norlandulo 2.0 could be a strong contender in the future.

Mindmap

Keywords

💡3D scanning

3D scanning is the process of creating a three-dimensional digital model of a real-world object or environment. It involves capturing data on the shape and color of the subject and then using that data to produce a 3D model. In the video, 3D scanning is the central theme, with a focus on the technological advancements that are making it more accessible and its various applications in industries like entertainment, medical, and engineering.

💡Photogrammetry

Photogrammetry is a technique used in 3D scanning that involves taking multiple photographs of an object or scene from different angles and then using software to stitch these images together into a 3D model. It relies on the presence of distinct features in the images, such as color changes or textures, to create a detailed mesh and texture quality. However, it may struggle with objects that lack these features, like single-colored or highly reflective surfaces.

💡Kiri Engine

Kiri Engine is a software application mentioned in the video that provides tools for 3D scanning and augmented reality (AR) technology. It offers multiple ways to create 3D models, including traditional photogrammetry and a new feature called 'featureless' that utilizes neural rendering techniques. Kiri Engine is available for both iPhone and Android devices and is being highlighted in the video for its innovative approach to 3D scanning.

💡Artificial Intelligence (AI)

Artificial Intelligence refers to the development of computer systems that can perform tasks typically requiring human intelligence, such as learning, reasoning, problem-solving, and perception. In the context of the video, AI is used in Kiri Engine's 'neurological' feature to reconstruct 3D scenes from images or videos, overcoming the limitations of traditional photogrammetry by not relying on distinct features in the images.

💡Augmented Reality (AR)

Augmented Reality is a technology that overlays digital information, such as images, videos, or 3D models, onto the real world, enhancing the user's perception of their surroundings. The video mentions the growing market for AR and its relation to 3D scanning, as the technologies become more accessible and integrated into various applications, including gaming, education, and design.

💡Nerf technology

Nerf, short for Neural Radiance Fields, is a machine learning-based rendering technique that uses artificial intelligence to generate 3D models from a collection of 2D images or a video. It can create detailed 3D scenes and objects, even from images that lack distinct features, which traditional photogrammetry might struggle with. Nerf technology is highlighted in the video as a significant advancement in the field of 3D scanning.

💡Neurological

Neurological, as used in the context of the video, refers to a technology that builds upon Nerf by not only generating neural-rendered scenes but also producing high-quality 3D models that are usable in 3D software like Blender. It represents an advancement over Nerf by addressing some of its limitations and offering a more viable solution for creating 3D models for various applications.

💡Orlando

Orlando is a research paper published by Nvidia that introduces a new method for creating 3D models using neural networks. It represents a significant advancement in the field of 3D scanning and modeling, as it can produce high-quality 3D models from images that traditional photogrammetry might find challenging. The technology from the Orlando paper is used in Kiri Engine's 'neurological' feature, making it a powerful tool for 3D scanning.

💡Point Cloud

A Point Cloud is a collection of data points that represent the external surface of an object in three-dimensional space. These points are used to describe the shape of the object and can be converted into a 3D model. In the context of the video, Point Cloud is generated through photogrammetry or neurological technology, and it is a crucial step in the 3D scanning process.

💡Mesh Quality

Mesh Quality refers to the structure and precision of a 3D model's surface, which is composed of a network of points, lines, and faces (the mesh). High-quality meshes have a uniform distribution of polygons and avoid artifacts like holes or excessive distortion. In the video, mesh quality is an important aspect when evaluating the effectiveness of different 3D scanning techniques, with photogrammetry typically providing good mesh quality but Nerf technology initially lacking in this area.

💡Texture

Texture in 3D modeling refers to the appearance of the surface of a 3D object, which includes color, pattern, and other visual details. It is a crucial component of 3D models as it gives the object a realistic look and feel. In the context of the video, texture is discussed in relation to how photogrammetry and neurological technologies capture and reproduce the surface details of the scanned objects.

Highlights

3D scanning is becoming increasingly popular with the rise of artificial intelligence and augmented reality.

Kiri Engine is a versatile tool for 3D scanning and AR technology, offering an alternative to popular iPhone 3D scanning apps.

Kiri Engine now supports a new feature called virtual scanning using NeRF technology, which is particularly interesting for those without the necessary hardware for real-time 3D scanning.

NeRF technology works by using AI to reconstruct 3D scenes from a set of images or a video, making it possible to create 3D models of objects that lack distinct features for photogrammetry.

Norlandulo is an advancement over NeRF, providing high-quality 3D models that are usable in software like Blender.

Kiri Engine's implementation of Norlandulo technology is a significant step forward in the accessibility of high-quality 3D scanning for both iPhone and Android users.

Photogrammetry, while effective for detailed geometry and texture capture, has limitations with objects that lack feature points such as color variations or textures.

NeRF and Norlandulo technologies aim to overcome the limitations of photogrammetry by not relying on feature points and using AI to generate coherent point clouds and 3D models.

Kiri Engine's featureless scanning mode allows users to scan objects with limited feature points, such as single-color or reflective surfaces, which were challenging for traditional photogrammetry methods.

To use Kiri Engine's featureless scanning, users record a video of the object from various angles and submit it to the cloud for processing.

Kiri Engine's featureless mode is based on the Norlandulo algorithm, which is designed to address the shortcomings of both NeRF and photogrammetry.

Kiri Engine offers additional features beyond the featureless mode, such as an extensive access library, in-app editing, and texture toning.

NeRF and Norlandulo technologies represent a significant advancement in the field of 3D scanning, with the potential to revolutionize various industries, including entertainment, medical, and engineering.

The Kiri Engine's development and implementation of NeRF and Norlandulo technologies demonstrate the rapid advancements in AI and machine learning within the 3D scanning field.

The Kiri Engine is a testament to the growing accessibility of advanced 3D scanning technologies, providing users with powerful tools regardless of their technical expertise.