Progress 4 (20/2/2012 - 26/2/2012)
After sorting out some problems regarding the motion tracking, we progressed on making the GUI for the applications, so that it'll be more accessible.
Then, we sorted out things like blogs, posters and such, as part of the requirements for the project. Finalization commenced on Friday in this aforementioned week.
Then, we sorted out things like blogs, posters and such, as part of the requirements for the project. Finalization commenced on Friday in this aforementioned week.
Progress 3 (13/2/2012 - 19/2/2012)
Next, we progressed to implementing motion tracking application by using MATLAB environment. The concept of doing this application is by breaking images into sequences, then initialize a set of tracking boundaries to dignify the moving object.
Acquisition for the images requires steps as pertained by MATLAB tutorial:
Acquisition for the images requires steps as pertained by MATLAB tutorial:
The device (i.e. webcam) is the primary input for this. Then, boundaries using the regionprop feature are made.
Progress 2 (6/2/2012 - 12/2/2012)
We've come out with codings for both of the object detection during this aforementioned week.
Notice that the boundary specified in the for command varies from one to another. The ratio command materializes the area on which the objects are detected.
Notice that the boundary specified in the for command varies from one to another. The ratio command materializes the area on which the objects are detected.
Progress 1 (30/1/2012 - 5/2/2012)
A few discussions were conducted in this whole week, between us group members, to determine the algorithm on our object detection. Finally, we came out with this flowchart:
We decided on conducting the detection on two sets of objects, a circle and an A4 sized paper. The algorithm for both of the project remains the same, since the only variable that changes is the regionprops function. Images are being captured by a webcam purchased specifically for this project.
We decided on conducting the detection on two sets of objects, a circle and an A4 sized paper. The algorithm for both of the project remains the same, since the only variable that changes is the regionprops function. Images are being captured by a webcam purchased specifically for this project.
Second Project Meeting (30 January 2012)
The second project meeting was held in the first day of the project week itself. In this meeting, we've talked about our studies and research regarding the ways on implementing this project in MATLAB.
Due to time constrain, we decided on just A4 sized paper for the main material in object detection, as the boundaries are easier to code and implement.
Website such as Mathworks proved to be very useful as it introduced us to the functions that we need for the object detection. But, it was still a bit complicated for us as our knowledge in programs and codings were insufficient.
However, we came across some functions called imread which calls up the image that we are going to use for the project. Then, another useful function which we were introduced to is the regionprops and this function basically sets the region size for our detection material. When it comes across materials that are similar to the dimension specified in an image, the code makes up a colored boundary around it.
Due to time constrain, we decided on just A4 sized paper for the main material in object detection, as the boundaries are easier to code and implement.
Website such as Mathworks proved to be very useful as it introduced us to the functions that we need for the object detection. But, it was still a bit complicated for us as our knowledge in programs and codings were insufficient.
However, we came across some functions called imread which calls up the image that we are going to use for the project. Then, another useful function which we were introduced to is the regionprops and this function basically sets the region size for our detection material. When it comes across materials that are similar to the dimension specified in an image, the code makes up a colored boundary around it.
Subscribe to:
Posts (Atom)