Comments by PGT - CodeProject

PGT 22-Oct-15 10:56am View

Well i use this in a vertical filter on 3d depth data for Robotic control.
points[] only exist in the filter area of the code, not globally outside

Its inside an x row calculations and its not allowed to be on x-n or x+n, strictly only the current x. As to compare x,y against x-n,y or x+n, y strict horizontal filters are used.

The whole routine would never be called to fast, of which i can be sure since the robot that i control simply cant move that fast. A next robotic task comes after 2400ms. Cuyrrently I am fast enough (my math takes 120ms maximum). My math wouldnt work that fast without parallelism, I depend a lot on parallel code; and preferably i'd love to make it even faster. Because then i could make the behaviour of the robot smarter with some extra math.
There for i wonder this.

PGT 22-Oct-15 2:38am View

Well the math never uses the same point[] data in another threat, that i am sure.
Its only that i wonder since its an array if each threat would lock the whole array to write its result too, or only lock the specific array element.

PGT 29-Oct-12 10:59am View

Reason for my vote of 5
great code samples

PGT 27-Oct-12 15:11pm View

on seccond thought this is to much, i think i have to split the task between 2 computers and 2 cams deviding the task in 2, still it need to be fast, but it would allow the code to do some more actions

PGT 27-Oct-12 15:03pm View

i need it to fastly count certain pixels that fall into a heu range (or RGB with a work around). needles to say each frame is a new picture and a new count

PGT 27-Oct-12 14:59pm View

ough.. that cpu calculation it does hurt.
I still had to buy a computer for it, maybe an intel octocore with dual threads (16 cpu cores) but as you mention these cores will do more then running my code.
maybe cuda is an optoin i dont know it.. hmm

i just tought maybe i could do away with hsl but check for ratio's in rgb collors
or make the setup so that i can use a channel for it. (this was not what i intend its a workaround)

PGT 27-Oct-12 12:18pm View

hm there's tons of information in there thanks.
is there a route that you would recommend me? (since there all new to me I got lots to learn)

PGT 27-Oct-12 8:45am View

well i didnt want to point people in certain directions at first, hoping for some new ideas, but i should have given that info your right, the image is RGB format , frame size 800x600 (its not usb based) and actually that's the low res mode.

And i've been thinking (i'm still in doubth) to do HSL, because that's more usable.
The processing is actually quite simple i need to detect and count how many dots are in a certain color range. (heu) So there is no facial detection or so, there are no shapes or 3d things, but on the backdraw i got to process every pixel, I cannt skip pixels. In contrast to face or 3d detection here the challenge is more basic while still though.

PGT 27-Oct-12 7:31am View

yeah i know getpixel/setpixel wont do the job i need fast direct access, maybe even multi-threaded so each thread does a part of the frame.. (but that gives some overhead too creating a multi threaded app).. or perhaps launch is frame processing in its own thread.. i'm not really sure the only thing its got to be extreme fast.
And thats what wories me because i havent found it yet. here each milliseconds count for processing an image.

PGT 27-Oct-12 7:26am View

i fully understand, and there is my problem too
eventually i will even need faster camera's so 120fps is just a starter..
thats why i'm desperate for extreme fast code

PGT 1-Oct-12 15:37pm View

If it has been done, could you tell me?.
I've seen a lot of HTPC programs that required a graphical screen.
But i like one without screen, so i wont have to put on my TV to listen to radio, or select a different MP3 song.
Do you know of any that doesnt require a screen ? and would except remote input ?
I dont know and was thinking of writing one from scratch indeed

I've been looking for it but couldnt find something like it.

PS
(for me it would be easier to write in python then c++, because i'm better at python
also python is a higher language so usually its les trouble to create something
the reason i dont choose python for it is that i want to learn from the excersize
too. However if it has been done.. well then there is less reason to make it.

PGT 15-Nov-11 4:31am View

Ah i got confused in multiple ways
the sort command made that tpc001 is actually element number 1 not zero
another thing that got me confused is that if you do
document.write(sumar.push[1]);) >> wil output 1
therefor i was thinking it would show its index number
but it doent it show the length which 1

thanks for your help i was looking to long at it; got confused

PGT 3-Nov-11 11:58am View

See my answer to AspDotNetDev;
I am not sure who to credit for this, because i tried so many solution before asking it here, and i only was in the wrong document mode for IE.
I had tried JSON.stringify before too

PGT 3-Nov-11 11:56am View

Okay.. stringify is i think for now the easiest method.
But there might be more methods, as one would think especially in sharepoint that there would be buildin functions.

Anyway, there is more...
Apparently JSON has been added to Internet Explorer since version 8.
And starting from i think IE7, one can define the document mode for IE browsers. So JSON only works correctly in Internet Explorer if the document IE mode is set to at least IE8
So to use JSON inside the header the following meta tag is required.

<meta http-equiv="X-UA-Compatible" content="IE=8" >

Firefox just works, doesnt care about it.
But this was the main reason why i couldnt handle the returned objects.

Comments by PGT (Top 14 by date)