Creating 3D images using a 2D camera

Walid Ibrahim, Mohammed Farrokhi Aghdam, Omar Al-Safi
supervised by Dr. Dimitri Androutsos

Currently, there are a few camera-equipped smart cellular devices on the market that can take 3D images. They do that by using two identical cameras that are mounted in close proximity to each other and are slightly angled sideways. Each camera takes a single image of the scene, thereby proving two slightly different views of the scene. The two images are then combined together by using software in order to find the common points between them and extracting the depth information before displaying a 3 D image on a 3D-eneabled display. The purpose of our project is to eliminate the need for two cameras and instead use smart phones that are equipped with only one camera to capture the two viewpoints of the visual scene. By doing it this way, a cost saving will be achieved on the hardware by reducing the number of required cameras to one without significantly affecting the quality of the resulting images. Additionally, the extra amount of software processing due to the hardware reduction was insignificant due to the resource-rich smart phones that we have nowadays with features like large amount of memory and multi-core processors. The chosen platform for the designed software application was Android due increasing popularity of this platform as well as the availability of software libraries that significantly reduced the development time.

Project targeted applications: producing 3D images using single-camera Android smart phones that have a 3D display.