Layered Architecture of BubbleBoy

            Layered architecture ( figure) of BubbleBoy is a detailed interpretation of main components of the system. Layers of the systems designed as tree tier architecture. Main tears can be identified as Presentation Tier, Logic Tier and Data Tier. It will clarify the system functionality in abstract and it will clear the main layers and sub layers of them.

Logic Tier

This layer concludes the software application part of the BubbleBoy.

·         Image Acquisition

            Preparing of the images taken from user and feeding them to image processing will be done at this sub layer.

·         Image Processing

            This layer consists of three parts 2D image processor, 3D Image generator, Head Direction capture module. Initially the image is in the position came from the image acquisition sub layer and it will contain the marker with it.

·         Gesture Recognition

            This represent the recognition of the pose of the user in one time instant and the details will be compared with the knowledge base data. Then the results will be proceeds to the action detection layer.

·         Action Detection

            In this layer it will map the sequence of the images processed by the above layer with the details on human actions in Knowledge base and will detect the action of the user.

·         Knowledge Base Access Handling

            This layer is responsible for handling the communication of Knowledge Base with Logic Tear components. It will control the access requests from application and manage the data flow of the in and out of Knowledge Base.
Environment Generation

            This layer will take inputs from the action detection layer; generate appropriate environment change according to detected action of the user.

·         Environment Modification

            Environment Modification layer is the layer that will make changes immediately after the changes in environment are identified by the Environment generation layer.

Data Tier

·         BubbleBoy Database

            Database will be required for the Logic Tier components to store data apart from Knowledge Base. The Knowledge Base will be keeping specialized data for gesture recognition and action detection. Therefore the database for the other components of the Logic Tier will be necessary. 

·         Knowledge Base

            Knowledge Base will consist of two types of data basically. Gesture model details and Action detection details can be identified in this stage as details to be included in Knowledge Base.

 Presentation Tier

·         Graphical Environment

            This will be showing the virtual environment to the user according to the action he/she presented.

Four Major Phases of BubbleBoy

The system BubbleBoy can be divided in to four major phases in its development process. They are,

                               I.            Human Computer Interaction phase

                            II.            Image Analysis Phase

                         III.            Action Recognition and

                         IV.            Environment

Phase   I

            This step is the initiation step of the complete operation. It will capture images and feed them to the phase II. It will capture the image with passive markers on user. Two cameras will be used to the capturing operation.

Phase II

            Initially the image is in the position came from the first phase of the system and it will contain the marker with it. It will progress in to next step. Then it will generate limb recognition with the help of the markers. The most critical part of this phase is in the next step. The 3D model will be generated using limbs recognized. Angles and the sizes will be main information for the conversion.

Phase III

            Since this need to be implemented for a series of images which are fed by video camera, it need to be evaluated all incoming images efficiently. So the series of generated 3D models will be supplied to the next phase. This phase’s main task is to match the action according to the trained set of actions.

Phase IV

            This will be the user feedback phase as the user can watch the results of the action he/she presented.

Project meeting minute - 09

Date & time, duration

The first meeting was held on 2010- 12 -02 for one hour

Venue

Office room of Dr. Chandana Gamage

Members present

All the members were present there.

Topic of the discussion

Feedback on the second iteration of the Software Architecture Document (SAD)

Brief Discussion of the discussion

Today also we basically discussed about the content of Software Architecture Document. This time we presented our second iteration of the document and sir explained us the deficiencies we have made there. Those are

• Maintain the consistency of the document control

• Come up with diagrams that looks fine

• Use correct terms such as gesture recognition, image acquisition and action detection.

• Design document consisting of two components such Final design of our proposed system § as Assurance of our design with the work carried by other people(taken § from trusted sources)

• In the image acquisition phase sometimes we have to exclude some of the gestures depending on the context where we will be at that particular moment (attention on gesture inclusion and gesture exclusion)

• Explain the design for the 3-D model creator and its state of art(While we have design for overall system we have to have designs for all the sub components come under the overall system)

• And for every design it should carry the rationale behind it, historical work, current trends and future directions.

• And also we were asked to look at the web resource “http://www.cs.sfu.ca/~mori/research/”

Key topics discussed

1. Compare the things we did with other people in the history who has contributed to the things we are referring to.

2. Maintain the standard of the referencing process.

3. Maintain a log book for the things we carry out with related our project

Comments and agreements

Chandana sir made us understand the important points that we have to improve and asked those corrections to be made in the next iteration of the document and send him that version of the document tomorrow. And since the references are not in standard way once they are created from different ways sir suggested us to go to the latex which uses an approach called bib Text to have references in standard way.