Natural manipulation on 3D object using fingertips markers in Augmented Reality System

Lai , Shi Yong (2013) Natural manipulation on 3D object using fingertips markers in Augmented Reality System. Universiti Malaysia Sabah. (Unpublished)

[img]PDF
1376Kb

Abstract

The trend of augmented reality has currently switches to the study of interactive augmented reality. The main goal of this project is to develop an interactive augmented reality system that enables user to manipulate or perform standard transformation on 3D virtual object by using fingertips markers. Fingertip casing is invented to fit the fingertip marker on top of fingers. New markers are generated as the system requires multiple marker patterns. Interaction techniques such as selection and manipulation techniques are applied in order to perform standard transformation on 3D virtual object. Computation of distance and angle between markers are integrated with the interaction applied to perform standard transformation. Experiments analysis is done by conducting tests on factors that affect the marker detection. From the result of light condition tests, observation concludes that initial markers detection for dim lighting test is in a range of 2 to 5 markers, while moderate lighting test detects markers in a range of 4 to 6 markers. Comparisons between dim lighting and moderate lighting tests show that dim lighting test has less potential to detect more markers, thus concludes that markers detection works better in adequate lighting supply. However, number of markers that detectable in motion is yet inconsistent, hence proves to be the limitation of ARToolKit. On the other hand, results of distance tests show that shorter distance between markers and camera yields better result of 6 markers detected while longer distance gives the result of a range of 5 to 6 markers. Conclusion of the tests is the farer the distance between markers and camera, the harder the markers to be detected. For frame rate test, the frame rate is 4 frames per second before interaction is applied to the system with all markers detected, and the rate goes up to 10 frames per second after interaction between fingertips markers begins. The result shows the average memory consumption of the program is 35.05 MB after 20 sets of experiments are done to calculate the average memory consumption of the program. In conclusion, future study can be focused on lighting condition that affects marker detection in ARToolKit as well as motion capture by ARToolKit in order to produce smoother interaction between user and the virtual world.

Item Type:Academic Exercise
Uncontrolled Keywords:Fingertip casing, ARToolKit , Hybrid Tracking Techniques , Interaction Techniques , Standard Transformation Performed
Subjects:Q Science > QA Mathematics > QA76 Computer software
Divisions:FACULTY > Faculty of Science and Natural Resources
ID Code:14816
Deposited By:IR Admin
Deposited On:15 Nov 2016 14:58
Last Modified:15 Nov 2016 14:58

Repository Staff Only: item control page


Browse Repository
Collection
   Articles
   Book
   Speeches
   Thesis
   UMS News
Search
Quick Search

   Latest Repository

Link to other Malaysia University Institutional Repository

Malaysia University Institutional Repository