Closed ElliotAOram closed 7 years ago
During development of this feature I have discovered that OpenCV in python does not operate in the way in which I expected. Python OpenCV uses 3 channel (r,g,b) 2D numpy matrices as the frame for output windows. This means several changes to the design have to be made to allow for this:
video_feed_array
is no longer required
copy_video
and populate_video_feed_array
are no longer required
The changes made in Feature 3 will be rolled back before work on this issue continues. As little work has been done this branch for this issue a new branch might be created from the reverted master.
I have deleted my local branch and the repo branch copy and will restart this issue at the design stage
The process_and_output_video
function will not be able to be tested too easily as it runs in an infinite loop that is only broken on user i/o. As such I have made the choice to refactor the functionality to a HelperFunction
package. This should mean that this functionality should be able to be tested even if the direct output is not
Added in helper functions:
int
parsers for checking parameter (required as python has no types)These functions will be used in the process_and_output_video
function to correctly place images
Due to further refactoring the design needs to be updated
Feature Description
The vpa application should be able to output the video feed to a window that can be used for display. Furthermore, this should be able to output four copies of the video feed to a single window. (The task of rotating the video feed will come after this, Feature 4)
Tasks