Chris Pollett > Students >
Sneh

    ( Print View)

    [Bio]

    [Blog]

    [CS 297 Proposal]

    [Deliverable 1]

    [Comparison between ARKit and ARCore - PDF]

    [2D to 3D Cartoon Modeling - PDF]

    [YOLO Object Detection iOS - PDF]

    [Deliverable 2]

    [Deliverable 3]

    [ARCritique - PDF]

    [Deliverable 4]

    [CS 297 Report - PDF]

    [CS 298 Proposal]

    [CS 298 Report - 3D AR Reconstruction - PDF]

    [3D AR Reconstruction Slides - PDF]

Building AR app to build 3D mesh from 2D image data

Description: The deliverable stared out with the idea of building a mesh using Lidar scanner on the iPhone. Eventually we got to know that the Lidar scanner is available only on the Pro models. Hence we shifted to an alternate stratergy. This was using the photometry API provided by Apple. I built two apps for this. The first one is used for capturing images from the phone. We built an app for this because we need depth and gravity information from the images hence it made sense to build an app. After capturing 100s of images we transfer them to a laptop where we use the photometry app on the MacBook to build a mesh. Photometry API's are only available on the MacBook hence we have to run the Mac App. Following this we get a usdz model which is the 3D mesh of the object.

Example: When the images are used by photometry API it builds a 3D mesh and outputs it in the usdz format.

Mesh of the basketball 3D mesh of the basketball

    [Deliverable 2- 2D images to 3D model - Zip]