My project aims to create an AR experience using around 200 image targets, each linked to video, 3D, or combined content; including the creation of an external database where each image target is associated with its corresponding content (video, 3D, or both) and unique identifying data, and categorized in areas.
The process should be something similar:
- Image Capture: The user scans a real-world image using their device’s camera.
- Area Identification: Machine learning or image recognition algorithms analyze the captured image to determine its associated area.
- Image Target Loading: The system loads the set of 8th Wall image targets specific to the identified area.
- Content Retrieval: The relevant video, 3D, or combined content, along with identifying data, is fetched from an external database based on the activated image targets.
- Content Display: The retrieved content is displayed in augmented reality, overlaid on the real-world image.
My question about this project is regarding performance (considering optimized 3D models and efficiently encoded videos). Can the loading of content from an external database be optimized, making the experience similar to an app with few image targets and content loaded only on that app?"