Nillan Nimal, Abhishek Menon, Daniel Velyvis, Mitchell Carlos-Carneiro, Sajad Saeedi

Toronto Metropolitan University

In recent years, 3D scanning has gained popularity due to its ability to create precise and accurate 3D models, which have applications in various fields such as engineering, design, healthcare, and entertainment. However, current desktop scanners are expensive, which limits their accessibility. To address this issue, we present a cost-effective and efficient desktop scanner that utilizes a regular smartphone to capture images of an object, which are then processed using photogrammetry algorithms to produce a 3D model. The scanner can generate high-resolution textured models that accurately represent the original object. The scanner's geometric accuracy was evaluated by comparing its results with a Structure Sensor Pro. The comparison demonstrated that the scanner could generate models with precise geometry.

Our prototype provides an efficient and low-cost means of automated desktop scanning by using a smartphone to capture images of an object. The prototype scanner incorporates a user-friendly touchscreen interface, which enables users to input essential parameters such as focal length, smartphone camera position relative to the device's center, smartphone thickness, object diameter, and object height.  These input parameters facilitate the precise positioning of both the smartphone camera and the object, allowing for the capture of high-quality images from four distinct viewing angles. Upon the completion of the scan, the images from the smartphone are retrieved and sent to AliceVision Meshroom for processing.

If you have any questions, feel free to reach out to us at the following email: nnimal@torontomu.ca