/KinectProject

Xiuyuan Chen

Primary LanguageC

Zen blocks

  • Xiuyuan Chen
  • 20027485
  • MSc Creative Computing
  • University of the Arts London: Creative Computing Institute
  • Supervisor: Phoenix Perry
  • Submission Date: November 19th, 2021

Introduction

image

Seldom think about philosophy spontaneously. People don’t really understand their id, ego and superego because of the lack of an appropriate environment to calm down and think. Normally, people rely on the way they look at themselves in the mirrors. However, the mirror only reflects peoples outer appearance, which draws too much attention on the looking. People rarely calm down and think about their inner hearts. The concept of the Zen garden helps people stay there to calm down, and find their inner peace. At the same time of interaction with people, the installation weakens the appearance of people, so as to cause people to think more about their own inner. Screens on the walls are the existing environment that the public spaces have. By creating neat blocks on the screen, the uniform shapes make people calm down. Adding a Kinect sensor to detect people, the wall can interact with people by transforming their outer shape. As a result, it starts meditation and helps people calm down and find their inner peace.

Basic Setup

Environment:

Iteration Process

Online communicate with tutor Date: 2021.7.20
Propose the topic with the tutor, establish the research direction, and make a proposal Supervisor feedback: the proposal lacks critical thinking and a level of intellectual inquiry :think about how this could be a more meaningful experience.

  • Midify the proposal
  • Create the research list
  • watch totorials related to Kinect

V1: Low resolution depth image

In order to fit most low-powered computers, my first thought was to reduce the amount of computation

V2: Removing the noise

I don’t need the background information. For this step, I use text() function to see the distance of every pixels, then I use if() statement to remove the high numbers.

V3: From 2D to 3D

The 2D canvas was size(512,424); Then I change the canvas to 3D size(512,424,P3D);

As a result, the the depth image is transformed into point cloud

V4.1: More attempts V4.2: More attempts

I also made some visual change of every pixels to make it looks more 3D-like

V5: Adjust the threshold

I use the MouseX and MouseY position to adjust the Minimum and maximum threshold of the depth image. I can adjust the threshold depending on the environment.

V6: Adding the boxes, Adding the velocity

Instead of rectangular shape, I create boxes to make it looks more realistic

Also I add velocity on every boxes to make them recover slowly

V7: Final adjustments

  • Size adjustment
  • Color adjustment
  • To put the project on the large screen, I add int width and int height to make it adjustable
  • Also, I make color changes to make it looks more comfortable

Thanks to