Current Server Time: 04:47 (Central Europe)

#apertus IRC Channel Logs

2019/04/21

Timezone: UTC


00:57
futarisIRCcloud
joined the channel
02:24
intrac_
joined the channel
02:27
intrac
left the channel
06:37
BAndiT1983|away
changed nick to: BAndiT1983
07:15
Nira|away
changed nick to: Nira
08:07
Nira
changed nick to: Nira|away
08:08
Nira|away
changed nick to: Nira
08:18
Nira
changed nick to: Nira|away
08:18
Nira|away
changed nick to: Nira
08:19
BAndiT1983
changed nick to: BAndiT1983|away
08:19
BAndiT1983|away
changed nick to: BAndiT1983
08:30
Nira
changed nick to: Nira|away
09:28
Nira|away
changed nick to: Nira
09:29
Nira
changed nick to: Nira|away
09:29
Nira|away
changed nick to: Nira
09:50
Nira
changed nick to: Nira|away
09:50
Nira|away
changed nick to: Nira
09:53
Nira
changed nick to: Nira|away
09:53
Nira|away
changed nick to: Nira
10:03
Nira
changed nick to: Nira|away
10:04
Nira|away
changed nick to: Nira
10:06
Bertl_zZ
changed nick to: Bertl
10:06
Bertl
morning folks!
10:12
Nira
changed nick to: Nira|away
10:13
Nira|away
changed nick to: Nira
10:37
Bertl
off for now ... bbl
10:37
Bertl
changed nick to: Bertl_oO
10:44
Nira
changed nick to: Nira|away
11:41
Bertl_oO
changed nick to: Bertl
11:41
Bertl
back ...
12:21
BAndiT1983
changed nick to: BAndiT1983|away
12:55
Ashu
joined the channel
13:25
Ashu
hi Bertl !!
13:26
Bertl
hey! how's it going?
13:26
Ashu
I read in lab T931,It is planned that data-points for histogram are planned to be send to web client as JSON data. Would we be using some code to parse overlay data(data-points) to json data for sending to web client?
13:27
Ashu
everthing is fine :)
13:28
Ashu
https://lab.apertus.org/T931
13:28
Bertl
no, the idea is to calculate the data from the image stored in DDR memory (as fast and as efficient as possible)
13:28
Ashu
if this is right,then I think we can use d3 library to do this.
13:28
Ashu
I have tried to make example to use d3 library to render histogram for 5-bitdepth rawfile for which I am sending json data .
13:28
Bertl
and the either generate an overlay from this data or alternatively send the data to a web client
13:29
Bertl
where it is processed and visualized on the client side
13:32
Ashu
no, the idea is to calculate the data from the image stored in DDR memory (as fast and as efficient as possible): yes the calculation part is done there only but after that data must be send to client which can be done as json
13:33
Bertl
json as data transport is fine, yes
13:33
Ashu
https://github.com/ASHOK17198/Render-SVG-Histogram/blob/master/index.html
13:33
Bertl
the important part is to get all the calculations and visualization (drawing) done by the client and not by the camera
13:36
Ashu
yes I understand that and I have done this in the example.
13:37
Ashu
after getting data in client as jason I am rendering it in browser as SVG.
13:37
Ashu
*JSON
13:42
Bertl
excellent
13:42
Ashu
left the channel
13:42
Ashu
joined the channel
13:51
Ashu
So, can we proceed with this approach for visualisation for tools
13:51
Ashu
?
13:52
Bertl
sounds good to me, but I can't test it ATM ... check with BAndiT1983|away when he's around
13:54
Ashu
okay I will be checking logs . Thanku :)
13:54
Ashu
left the channel
13:56
Bertl
no problem!
14:07
micky
joined the channel
14:08
micky
left the channel
14:41
comradekingu
joined the channel
15:30
aombk
joined the channel
16:01
vup2
Will transferring the raw data and calculating the histogram completely on the client side be actually fast enough?
16:02
Bertl
not the raw data :)
16:02
vup2
But is that not what Ashu is suggesting?
16:03
vup2
(Atleast thats what is done by the file he linked)
16:03
Bertl
well, that's why I tried to explain the process: userspace tools collect 'data' about the image, e.g. the histogram buckets
16:04
Bertl
then this data is either used to render the overlay or is sent to a web client which then processes the data and draws the histogram
16:05
Bertl
the file he linked does draw histogram bars based on data
16:05
Bertl
(at least as far as I can tell)
16:06
Bertl
I've no idea where the data comes from or how it is generated
16:09
vup2
Yeah, ok he calculates the histogram buckets here https://github.com/ASHOK17198/Render-SVG-Histogram/blob/ed7ef697fa2b0898b9d41fa16164d431f78f3ca4/index.html#L81
16:09
vup2
In the webclient itself
16:10
Bertl
well, it is an example, no?
16:11
Ashu
joined the channel
16:13
vup2
Yeah but it sounded like they he wanted to do it the same way for the actual histogram
16:16
Ashu
Bertl: I've no idea where the data comes from or how it is generated : I have taken values of bucket which were calculated in challenge task T1140.
16:17
Bertl
as I said, the visualization per-se looks fine to me, i.e. that's the way to go
16:17
Bertl
the data, e.g. the bucket counts, need to be calculated like in the challenge task
16:18
Ashu
vup2: Will transferring the raw data and calculating the histogram completely on the client side be actually fast enough? : I am not suggesting to transfer whole raw data, I am suggesting to send only the calculated bucket values by hist.c
16:18
Bertl
and then can be either packaged in the generated html or sent via a side channel to the client code
16:18
vup2
Ah, sorry for the misunderstanding
16:20
Ashu
the data I have written inside json is hardcoded by me using the output I recieved from challenge task .
17:38
Ashu
left the channel
17:59
aombk2
joined the channel
18:02
aombk
left the channel
18:45
se6astian|away
changed nick to: se6astian
19:26
Dev_
joined the channel
19:26
Dev_
left the channel
19:26
BAndiT1983|away
changed nick to: BAndiT1983
19:26
Dev_
joined the channel
19:27
Dev_
left the channel
19:46
intrac_
left the channel
19:47
intrac_
joined the channel
20:58
se6astian
changed nick to: se6astian|away
22:44
intrac_
left the channel
22:45
intrac_
joined the channel
22:56
Lucky_
joined the channel
22:56
Lucky_
Join
22:57
Lucky_
left the channel
23:11
BAndiT1983
changed nick to: BAndiT1983|away