[453_tmc] Re: Camera Node Data
- From: rodger4@xxxxxxxxxxx
- To: 453_tmc@xxxxxxxxxxxxx
- Date: Thu, 2 Oct 2008 11:58:15 -0400 (EDT)
I don't think we'll need a table.
Actually, we may still use your boxes in an invisible way (just to draw
over the boxes). However, the more efficient way, since the boxes are
already in the images, is to write a function that asks the database if a
car exists at the point only when a user clicks on it.
This way, we don't have to use server CPU time querying the database for
the points and user CPU time drawing the boxes every time we update the
image.
I imagine that it will work something like:
- I update the feed with the most recent image for a camera.
- If the user clicks on the image, we us AJAX to see if the coordinates
the user clicked correspond to a car id by:
-- writing a function that requests a page, such as isthisacar.php, where
we pass controls x and y
-- when the user clicks on the image, say at (x,y)=(32,12), we call the
AJAX function to request isthisacar.php?x=32&y=12
-- the .php page just prints that car id or NULL if nothing is there
-- then we set "currently tracked car" variable = car id
Also, because we already have small-feed-click-to-main-feed functionality,
we don't have to call another function to bring that camera up-top.
I am sorry about them taking your job. However, if this works correctly,
it should be a more efficient and easier-to-implement solution (maybe).
It doesn't have to be done this way, of course. If we do go this way, is
this something that you'll want to do? If so, I can help. If not, I can
do it, and you can move to network API with Logen (again, if you want).
Also, let me know if you still want to meet today. I'm not sure if it's
still necessary if we don't use your stuff (again, sorry, I know you spent
a lot of time and put a lot of effort into it).
> The boxes that the camera group provides will not be able to be clicked,
> to change the camera and track. So would we resort to making a table for
> the user to select a car?
>
> -Dan
>
>> That's an excellent question. You could help Logen with the Network
>> API.
>> I think it's going to be a big job, so two minds should help. You guys
>> are going to have to write a program that interfaces with our database
>> and
>> the network, so you could split the job that way (i.e. one of you take
>> database interfacing and networking for the other). The network guy is
>> going to have to write code that decodes the buffers and uses the
>> database
>> guy's functions to write to the database. The database guy is going to
>> have to periodically check the database for new commands and use the
>> network guy's functions to send those to the cams (I think, unless Logen
>> decided not to).
>>
>> What does anyone think?
>>
>> On Wed, 01 Oct 2008 07:58:22 -0400, <dcmille@xxxxxxxxxxx> wrote:
>>
>>> So, apparently the Camera group is sending us the video with the boxes
>>> already around the cars. What exactly am I going to be working on now?
>>>
>>> -Dan
>>>
>>
>>
>>
>>
>
>
>
Other related posts:
