Ihsan Rabbi

Factors affecting the design and tracking of ARToolKit markers

The quality of ARToolKit markers plays a vital role in the performance of Augmented Reality applications, but currently there is no algorithm or quantitative measure to guide users for designing high quality markers and their reliable tracking. This paper presents eleven factors that are important for designing and tracking ARToolKit markers. The effect of each factor on the quality (accuracy, detection speed, and inter-marker confusion) of marker tracking system is investigated and the optimal value(s) of these factors are found. Using the optimal values of these factors one can achieve the goal of reliable and robust tracking of fiducial markers.

Indoor vision-based auditory assistance for blind people in semi controlled environments

This paper presents an indoor auditory navigation system for blind and visually impaired people using computer vision based approach. In our approach, fiducial markers augmented with audio information are placed in the environment. The blind user holds a webcam attached to the system in his/her hand and navigates through the environment. The audio assistance is provided to the user whenever a particular marker is detected by the camera and thus enables him/her to independently navigate in the environment. The proposed navigation system provides two types of guidance (1) free mode guidance (2) targeted mode guidance. In the free mode guidance, the user navigates freely in the prepared environment and the system informs him/her about his/her current position through the audio information about the surrounding based on the detected marker. In the targeted mode guidance the user is guided from a source (current) position to his/her desired destination in the environment through the shortest path. Five blind users evaluated the system (both modes) in a semi controlled environment. The results obtained from the experiments indicate that the proposed system is successful in assisting blind users inside a huge and complex building. In addition the solution is low cost and gives high accuracy rate.

The Effect of Multimodal Virtual Chemistry Laboratory on Students’ Learning Improvement

This paper presents a novel Multimodal Virtual laboratory (MMVL) for the learning of chemistry experiments. MMVL is a Virtual Reality environment where the user can perform chemistry experiments like a real world chemistry lab. The user can easily interact with MMVL through 3D interaction interface. The audio and visual information about each chemical objects are provided to its users. The system improves the learning capabilities of students in chemistry education. The analysis shows that the average learning of untrained student is 32.7% while that of trained students increased to 83.5%. Experiments reveal that confidence level in practical field of students who got training in MMVL is much better than those who did not use it.

Measuring the Student’s Success Rate Using a Constraint Based Multi-modal Virtual Assembly Environment

Personnel Training is considered as the most important prerequisite in the assembly operations of any kind of equipment/apparatus ranging from simple nut-bolt assembly to complex equipment (e.g., aircraft engine) assembly. This paper presents a novel Virtual Reality Training System (VRTS) for the constraint based assembly of a 3phase step down transformer. The ARToolKit [1] markers are used for interaction with the VRTS. The system improves the technical skills of students in the real assembly environment. The analysis shows that the average success rate of untrained students is 35.7% while that of trained students increased to 81.5%.

The applications of marker-based tracking in semi-controlled environment

Tracking the environment is the key challenge for the development of Augmented Reality (AR) applications. Marker-based tracking is most widely used technique in AR applications. This technique provides accurate, fast and robust tracking in controlled environments. On the other hand its performance degrades in semi-controlled environments due to lighting condition variation. To overcome these challenges and enhance the usability of marker-based tracking to semi-controlled environments, the concept of multiple pattern files is proposed. Multiple patterns of one marker are captured during different lighting conditions and kept in database for tracking. The proposed technique improves the marker tracking accuracy under varying brightness and contrast values. The method is compared with conventional technique of single pattern file by developing an augmented reality navigation system inside a complex building. The accuracy of conventional method is 77.36% that is increased to 95.27% by using the proposed technique.

Optimality of black to white ratio and information complexity for robust marker recognition

Fiducial markers are widely used in many Augmented Reality (AR) applications such as to manipulate AR objects, robot navigation, education and AR based games. ARToolKit is an open source, easy to configure and well-documented tracking system which is widely used for designing marker based AR applications. The quality of ARToolKit markers plays a vital role in the performance of these AR applications, but currently there is no algorithm nor quantitative measure to guide users for designing high quality markers in order to achieve reliable tracking. In this paper we experimentally analyze the effect of the two markers design attributes, including black to white ratio (B/W) and information complexity on marker recognition. Based on above analysis, we developed threshold based algorithms for optimizing the two attributes. These algorithms also provide various cognitive aids and helps to the users. The work done will help ARToolKit users to design high quality markers with little efforts and as a result will increase the efficiency and performance of their AR applications.

Analysing the attributes of fiducial markers for robust tracking in augmented reality applications

Tracking the position and orientation pose of camera is a critical challenge for different modern applications like augmented reality, robot navigation, robot localisation, and 3D modelling and surveillance systems. Marker-based tracking is the most active technique used for camera pose estimation. For the development of augmented reality applications different marker-based tracking toolkits are available that consists of specific set of fiducial markers. In this paper, various fiducial marker attributes are analysed that helps to increase the accuracy of marker-based tracking in augmented reality applications. Experimental modules are developed to calculate the optimal values for each attribute. The experiments are designed to analyse the marker size, distance between marker and camera, the marker speed along all axis, the environmental brightness, the lighting contrast in the environment and dependency of marker size on tracking distance. Experimental study shows that these attributes affect the marker tracking. Augmented reality researchers can use these findings for the development of more reliable and accurate application.

Sharp-Edged, De-noised, and Distinct (SDD) Marker Creation for ARToolKit

Fiducial markers are widely used in many Augmented Reality (AR) applications such as to manipulate AR objects, robot navigation, education and AR based games. ARToolKit is an open source, easy to configure and well-documented tracking system which is widely used for designing marker based AR applications. The quality of ARToolKit markers plays a vital role in the performance of these AR applications, but currently there is neither algorithm nor quantitative measure to guide users for designing high quality markers in order to achieve their reliable tracking. In this paper we studied the effect of edge sharpness, noise and markers distinction on markers reliability. We also developed the specialized algorithms for designing Sharp-edged, De-noised and Distinct (SDD) markers, which have low inter-marker confusion and are highly reliable.

Classification of Markers in the ARTool Kit Library to Reduce Inter-marker Confusion

Fiducial markers are widely used in many Augmented Reality (AR) applications such as to manipulate AR objects, robot navigation, education and AR based games. ARTool Kit is an open source, easy to configure and well-documented tracking system which is widely used for designing marker based AR applications. The patterns inside the markers often cause inter-marker confusion in the library. Currently there is no specialized algorithm to ensure the markers distinction in the library and to reduce inter-marker confusion. In this paper we present algorithms for classification of the markers in the ARTool Kit library in a given number of distinct classes. The degrees of similarities among the markers are computed and stored in an N-by-N matrix and then kmeans algorithm is used to classify the matrix into k classes. It is recommended to use markers from different classes to reduce inter-marker confusion. The marker from a class of smaller size will have minimal confusion with others.