Purpose

Object Scanner is a portion of my thesis project, written in C++, Objective-C and Swift.

The whole purpose of this project is exploring the GPGPU and graphics programming on Metal Api.

Summary

The following steps are followed in this project.

• Calibrating the rgb-depth frame and creating point cloud by using Metal compute shader.
• Creating the transformation matrix on consecutive frames (No error management.)
• Presenting the consecutive point clouds on the screen by using the Metal vertex and fragment shaders.

Acknowledgements

This project is portion of my Master Thesis project, you can check the whole project in here.

This project is developed by using the following articles/projects and the Computer Vision algorithms that I used in the master thesis.

• Kinect Camera Calibration
• MetalByExample
• Apple Matel Docs
• iOS-OpenCV-FaceRec

Supported OS & SDK Versions

• Supported build target - iOS 9.0, Xcode 8.1
• iOS Devices with A7 or higher chip. (As you may know, Metal doesn't work on simulators)

Data Types

This project simulates the and RGB-Depth camera and provides and UIImage as an rgb frame and a float array as an depth frame. (Array indices are corresponds to [y * image.width + x]) Test data can be found under the !TestData directory.