China
Africa
Explore chapters and articles related to this topic
Improving the visual quality of signboard design through a ‘signboard aided-design system’ using augmented reality
Published in Artde D.K.T. Lam, Stephen D. Prior, Siu-Tsen Shen, Sheng-Joue Young, Liang-Wen Ji, Engineering Innovation and Design, 2019
In order to develop the system effectively and efficiently, we used the game engine Unity for realtime rendering and an ARCore (Lanham, 2018) SDK AR application on a mobile device. The fully assembled system was then created to illustrate vision-based marker less tracking in the AR application to use in the aided design of signboards. The overall architecture of the signboard aided-design system is shown in Figure 4. Develop the signboard display system with AR marker less tracking: Google developed ARCore for marker less tracking, which uses the visual-inertial odometry (VIO) to allow motion tracking. VIO is an algorithm that combines visual simultaneous localization and mapping (VSLAM) information from the camera and measurements from the inertial measurement unit (IMU). ARCore allows us to track position changes by identifying and tracking visual feature points from the device’s camera image. After identifying an object’s position and orientation in 3D, we use the pose to help camera project or flatten the 3D virtual objects into a 2D image, which is then displayed on the device. With these functions, the AR system has been developed using the following four stages: (a) motion tracking; (b) environmental understanding; (c) creating a scalable and rotatable virtual object; (d) creating AR measurement.Construct virtual signboard modeling: We use the DirectX 11 Tessellation shader technology to simulate 3D identity systems which combine logo and text. The technology can take time and reduce the inconvenience caused by modeling complicated 3D text models. The designer can display a 3D virtual signboard conveniently without any 3D design experience by simply employing the map and scaling adjustment.Create storage for signboard images and a programming system: The system manages downloadable content and the local storage of images for saving files in a mobile format. Users can store their work in their own data storage (e.g. Imgur) and add a file path into the mobile app’s storage directly. The pipeline of creating signboard image storage is shown in Figure 5.Develop an interactive UI design: The reusability, download-ability and scalability of the components are important features for enhancing the functions of this application. To allow input operation and control of the mobile app from the user’s end, the interface design for the AR signboard display includes a main page, AR demo interface, signboard input data interface and user guide interface. The result of developing an interactive UI design is shown in Figure 6.
Three-dimensional CityGML building models in mobile augmented reality: a smartphone-based pose tracking system
Published in International Journal of Digital Earth, 2021
Christoph Blut, Jörg Blankenbach
Alternatively, VIO, as, for example, in ARCore, could be effective. As described by Google (10 December 2019, https://developers.google.com/ar/discover/concepts), ARCore finds unambiguously re-identifiable image features and tracks these across subsequent images, deriving a relative camera motion. The resulting poses are fused with IMU data. Additionally, ARCore uses the sets of detected features to find and derive planes, such as wall-, ceiling-, floor- or table surfaces. These planes could be utilized for a geometry matching process with the corresponding geo-referenced CityGML surfaces, to obtain poses in a global reference frame. Disadvantageous is that the plane detection relies on image features, so that flat surfaces must be sufficiently texturized. White walls, therefore, may not be detectable for example.