Autodesk Forma’s Post

Revolutionizing Underwater Asset Management with 3D Models and Point Clouds 🌊✨ In the maritime industry, 3D models and point clouds are transforming how we interact with underwater assets, paving the way for enhanced optimization, planning, and innovation. 💻 Understanding 3D Models: A 3D model is a detailed digital representation of a structure, built from numerous polygons. These models offer superior visualization, simulation, and analysis capabilities. Advanced algorithms reconstruct the shape of underwater structures from point cloud data, resulting in accurate polygon meshes. 📍 The Power of Point Clouds: Point clouds consist of 3D points that accurately depict the surface geometry of underwater assets, enabling: Precise measurements Detailed representations for better decision-making 🔎 Why Choose Sentinus? With clear footage captured from 8 cameras, Sentinus is perfectly equipped to generate high-quality point clouds and 3D models. Interested in seeing these innovations in action? Check out the 3D models linked in the comments below! ⬇️ For more information, contact us at info@blueatlasrobotics.com. #3DModels #PointCloud #UnderwaterTechnology #Maritime #Innovation #Robotics

3D models and Point Clouds is transforming how we interact with underwater assets, unlocking possibilities for optimization, planning, and innovation in the maritime industries.🌟💡 💻3D model is a refined digital representation of a structure, consisting of a large number of polygons. ◾Providing enhanced visualization, simulation, and analysis capabilities. ◾Algorithms reconstruct your underwater structure shape from point cloud data, resulting in a polygon mesh that accurately represent an asset as a 3D model. ◾Point clouds are sets of 3D points that precisely depict the surface geometry of underwater assets. ◾Enables the creation of precise 3D models, allowing for accurate measurements and detailed representations of underwater structures. 🔎 Sentinus captures clear footage from 8 cameras, perfect for generating point clouds and 3D models. #pointcloud #3dmodels #underwater #robot Click the link below in comment section to check out the 3D models ⬇️ Contact us at info@blueatlasrobotics.com for more info.

CityGaussian: Real-time High-quality Large-Scale Scene Rendering with Gaussians 🎯 ⠀ 🔍 Key Highlights: ⠀ - Innovative 3D Scene Reconstruction: CityGaussian (CityGS) leverages 3D Gaussian Splatting (3DGS) to achieve high-quality visual effects with real-time rendering speed. - Novel Approach: Utilizes a divide-and-conquer training method and Level-of-Detail (LoD) strategy, partitioning scenes into blocks for efficient, parallel training. - Enhanced Rendering: By compressing Gaussian primitives and selecting detail levels based on proximity, CityGS ensures fast, high-fidelity rendering across various scales. - State-of-the-Art Results: Demonstrated superior performance in large-scale scenes, such as a 2.7km² city area, capturing rich details and improving rendering quality. 🌟 Additional Insights: ⠀ - Efficiency: Overcomes GPU memory limitations by training scene blocks independently and using a global Gaussian prior for seamless fusion. - Real-Time Performance: The block-wise LoD strategy minimizes computation, maintaining smooth, real-time rendering even for complex scenes. - Wide Applications: Ideal for AR/VR, smart city modeling, aerial surveying, and autonomous driving, providing high-fidelity, real-time visualizations. 🔗 Read the Full Paper: https://lnkd.in/eNxVcBTu 🌐 Project Page: https://lnkd.in/eveJYPG6 ⠀ 🎥 Demo Video: Check out the amazing demo video! 🔊🔊

🌟 Sneak Peek Friday: Exploring 3D Environment & Object Properties! 🎥🌳 In today's edition of Sneak Peek Friday, we delve into the intricate world of 3D environments and objects, showcasing their dynamic features and underlying properties to build up virtual worlds for perception sensor simulation. Our short video offers a glimpse into five key aspects of 3D objects that should be validated before generating synthetic sensor data:   1. Default Appearance: Witness the initial appearance of objects as it is visualized for the human eye. 2. Primitive ID: Explore primitives that build up the virtual world. This visualization shows the triangles of each 3D mesh in a colorful way. 3. Geometry ID: Identify the specific geometry sub-components that are composed to objects. E.g. the doors, wheels and windows of the cars. The decomposition of the objects into these sub-components allows to assign different materials per object and to move the sub-structures independently. 4. OpenMATERIAL Existence: Discover if an OpenMATERIAL file exists for each geometry sub-component, highlighted in green for "yes" and red/black for "no." 5. OpenMATERIAL ID: Uncover the specific OpenMATERIAL IDs associated with different materials. This shows where the same material is used on different objects, like for all the windows on each virtual vehicle.    🎬 Watch the Video Below! It provides a captivating insight into the intricate details and properties of 3D objects and the five different key aspects. Btw.: The software in the video is our in-house tool for visualization and inspection of 3D objects, as well as synthetic and real perception sensor data on detection level like lidar point clouds that you might have seen in our former posts, as well 😉 So get in touch, if you would get an introduction to it!   🌐 Advancing Realism in Simulation: With ASAM OpenMATERIAL emerging as a vital standard, we're paving the way for enhanced realism and standardization in simulation environments. We build up the scenegraph for our sensor simulations based on this standard, ensuring compatibility and consistency across platforms.   Stay tuned for more exciting developments as we continue to push the boundaries of simulation technology. Join us every Friday for fresh insights and innovations!   #SneakPeekFridays #3DObjects #OpenMATERIAL #Perception #Sensor #Simulation and #Model #Validation

I create a stealth airplane concept in Creo Parametric & used Real-time rendering in KeyShot to see what the airplane would look like in the sky. I then asked AI to take my concept and show it in the sky above Miami, Florida. It took my concept and made some modifications to it based on real airplane data that it was familiar with. #aerospace #aerospacedesign #aircraftdesign #creoparametric #3dmodeling #3drendering #conceptdesign #Industrialdesign #ai #aidesign

3D models and Point Clouds is transforming how we interact with underwater assets, unlocking possibilities for optimization, planning, and innovation in the maritime industries.🌟💡 💻3D model is a refined digital representation of a structure, consisting of a large number of polygons. ◾Providing enhanced visualization, simulation, and analysis capabilities. ◾Algorithms reconstruct your underwater structure shape from point cloud data, resulting in a polygon mesh that accurately represent an asset as a 3D model. ◾Point clouds are sets of 3D points that precisely depict the surface geometry of underwater assets. ◾Enables the creation of precise 3D models, allowing for accurate measurements and detailed representations of underwater structures. 🔎 Sentinus captures clear footage from 8 cameras, perfect for generating point clouds and 3D models. #pointcloud #3dmodels #underwater #robot Click the link below in comment section to check out the 3D models ⬇️ Contact us at info@blueatlasrobotics.com for more info.

SIMULIA redesigned this bike's frame with smaller bars than on our previous model to reduce frame weight. 🚲 Check out how they did it. ⬇️ This electric bike designed for urban commuting leveraged a unified modeling and simulation approach in 3DEXPERIENCE Works with the collaborative cloud-based 3DEXPERIENCE platform.

3D Packing for Self-Supervised Monocular Depth Estimation 🛣🔥 Many robotic platforms typically rely on active sensors like LiDAR for direct 3D perception but monocular camera depth estimation is becoming better and better and you can get very high detailed and accurate depth maps. The next step is to pack the RGB images and depth maps into a point cloud which can be used for 3D perception. In this paper, they propose a novel self-supervised monocular depth estimation method combining geometry with a new deep network, PackNet, learned only from unlabeled monocular videos. A newer alternative could be DepthAnythingV2. The architecture leverages symmetrical packing and unpacking blocks to jointly learn to compress and decompress detail-preserving representations using 3D convolutions. 👉 Check out the GitHub Repo, paper and documentation here: https://lnkd.in/dmHQP-36 👉 If you want to start with freelancing, take your career to the next level or land jobs in AI/ML: https://lnkd.in/dRm7xC6K

SIMULIA redesigned this bike's frame with smaller bars than on our previous model to reduce frame weight. 🚲 Check out how they did it. ⬇️ This electric bike designed for urban commuting leveraged a unified modeling and simulation approach in 3DEXPERIENCE Works with the collaborative cloud-based 3DEXPERIENCE platform.

Introducing Spyder Spyder is a tool for capturing detailed store layouts and creating 3D digital twins of physical spaces. How It Solves the Problem: •Reduces the need for costly physical prototypes by enabling virtual testing and experimentation. •Captures precise data on store fixtures, assets, and layouts to support accurate planning. Key Features: •High-resolution photo walkthroughs of stores and Lidar point cloud scanning for spatial accuracy to create a 3D digital twin. Key Benefits: • It cuts the time and costs associated with physical store surveys and prototypes and enables rapid testing of new layouts and concepts. Mark Edwards Matt Martin Tom du Val Think Blue