Y2 project week 3: Sandbox Setup, Game and UI Test

Week 3: 

Sandbox Setup, Game and UI Test

In this week, tripod was used by the developer to setup Kinect and projector to test the sandbox. Besides, some versions of user interface for touchscreen have been finished and tested. Some problems of code efficiency has been solved. In addition, some updates of the sandbox game have been finished.

1- Group meeting (12/02/2019)

In the third meeting of our group, the following points have been discussed about..

-The user interface.

-Poster and its contents.

-The problems faced in the second week and how to solve it.
-Week 3 work.

2- Supervisor meeting (13/02/2019):

In the third meeting with the supervisor the following points have been discussed about..

-Some ideas about 3D printer to print sandbox surface.
-Some functions could be added like water motion.
-The lunar theme should be light and the shadow could be added.
-The poster

3-Lab (01/02/2019) 

Sandbox Updates

In this week, tripod was used by the developer to hold Kinect and projector to test the sandbox model, games and user interface. Figures and video below shows the result.

Figure 1:  Test the sandbox model

Figure 2: Test the lunar sandbox theme

                                                                                    Video 1:  Sandbox video

                                                           

User Interface Design

The aim of designing an user interface for the Sandbox is to let user easy to interact with the game itself. Last week, the touchscreen has been proved that is possible to be used as a carrier to run sandbox game and the user interface (UI) at the same time. Thus, in this week, the first and second versions of UI, which can be run on the touchscreen, has been designed.

Figure 3: First version of user interface on touchscreen

Figure 4: Second version of user interface on touchscreen

The UI graphic was drawn by using c++ openframework graphic toolkit (ref:https://openframeworks.cc/documentation/graphics/), and the button, which can be touched by the user was generated by using c++ ofxDatGui toolkit (ref: https://github.com/thomwolf/ofxDatGui). Figures below shows the code of UI graphic and buttons the group programmed.

screenshot 1:Add button code

screenshot 2: UI code

Code Efficiency Problem

However, it was found that drawing the user interface needs to use a lot of graphic processing units, which will reduced the response time of sandbox game. Thus, to make the program more effective and reliable, many methods were tried by the group members, and some solutions were found. The first method is to delete the pictures used in the UI background and draw the background using basic lines, this method can double the performance of the sandbox game.

Besides, it was found that loading different words font format will reduced the response time of the sandbox game. Thus, after discussing with members, considering the aimed users are most children, the solution was to use sketching to replace the words. Figure below shows the designer is designing the UI for the user.

Figure 5: The designer is designing the UI for the aimed user

"Make a Shape" Game Updates

The similarity check algorithm and real-time scroe display parts have already been done in this week. However, It was found that the user interface of "Make a Shape Game" is not very easy to understand by the users, after discussing with members, one solution was to print the shape that the user need to make on the sand, before the game started. Figure below shows this solution.

Figure 6: Print the shape that the user need to make on the sand at the start

Besides, another problem was on the real-time score display parts, to display real-time score, the similarity check algorithm will be called each frame, which will slow down the respon time of the whole program. Thus, the solution is to call the similarity check algorithm every three or more seconds, and it shows that this method improve efficiency significantly.

graph 7: Similarity check by sand

Setup Kinect and projector

In this week, the suitable distances to put the new projector and Kinect camera have been calculated. However, time is needed to build the holder for the devices. Thus, for now, tripod was used by the developer to hold Kinect and projector to test the sand box model. Besides, sand was used for real test.

Figure 8: Fill up box with sand

Figure 9: Tripod was used to hold Kinect and projector to test the sand box model

Water Flood Function Test

This video below shows the water flood function, after user touching the start button, the water level will increase slowly.

Video 2: The water flood function

Poster Updates

After discussing with the other group, we have decided to talk about the design of two posters next Wednesday. Besides, two groups want to try to make two posters joinable with each other. The following figure shows the first prototype of the poster.

Figure 9: The poster prototype

4-Problems,Issues and Concerns

-As mentioned above, if the user interface is used, it will cost lots of GPU and make the picture flash slow.

-The response time of real-time score display is slow.

-Touchscreen response time is slow.
-The feedback of touch screen button is hard to use.
-The sandbox two sides still has empty because the lack of holder.

5-Plan for next week

-Meet on Thursday and summary this week work.

-Design a comprehensive and brief poster.

-Try to solve the problems for week 3.
-Think about the report of project