Import divelog from camera

[Paul Buxton]

I suspect adding support for the format to subsurface is more effort than
it is worth, as I can't imagine there are many divers who are willing to
spend $750 on a dive camera that haven't already got a computer.

@Robert , the problem with that solution is it requires a an RGBD image,
where D is the distance to objects in the scene. I came across this
algorithm nikolajbech/underwater-image-color-correction: The algorithm can
color correct underwater images (github.com)
<https://github.com/nikolajbech/underwater-image-color-correction#readme> which
looks pretty simple to implement. By coincidence it appears to have been
developed by someone at Paralenz, but it doesn't have any knowledge of
depth. You can try it online Paralenz Color Correction - BETA
<https://colorcorrection.firebaseapp.com/> Testing it on a couple of still
images, it seems to produce pleasing results.
I will have a look at implementing it as a ffmpeg filter (there are a
couple of other candidate algorithms to try as well).

Paul

On Mon, Jul 19, 2021 at 6:04 PM Miika Turkia <miika.turkia at gmail.com> wrote:

> Looks like the time format is not currently supported.
> - The date uses different separator than what we support (would be easy to
> add)
> - Date and time are in same column, so would require special parsing to
> split them (should be easy enough)
> - Time is given in clock time, not relative to start of the dive. So this
> would also need to be taken into account. And this would probably be
> tricky, see below.
>
> Support could be implemented either by preprocessing, or adding support
> directly to Subsurface, just as you suggest. The latter is a bit more
> complex as the import is unfortunately done with XSLT and CSV import with
> it is really already pushing the limits of that language (it was meant for
> XML transformations). Would require a bit of thinking how to do the direct
> import, there might be some trick that I cannot think right now.
>
> The + sign Robet mentions would be easy to take into account, if it is not
> already handled correctly.
>
> miika
>
> On Mon, Jul 19, 2021 at 7:37 PM Paul Buxton via subsurface <
> subsurface at subsurface-divelog.org> wrote:
>
>> I may not have been very clear. I meant that the Paralenz might be an
>> alternate way of logging your dives, rather than using subsurface as a way
>> of getting image correction based on depth.
>> One of the problems with that is exactly what you asked for, i.e. having
>> images with associated depth data. Ideally you would want images with a
>> color reference chart.
>>
>> Paralenz have been good enough to send a sample csv file. Which it looks
>> like it would be easy enough to either pre-process to make it consumable by
>> subsurface, or add support for.
>> The fields are not quite a match for the ones in the csv importer, but
>> could be handled easily enough I think.
>>
>> Paul
>>
>> On Mon, Jul 19, 2021 at 10:29 AM Robert.Helling <robert at neu.atdotde.de>
>> wrote:
>>
>>> Hi,
>>>
>>> On 19. Jul 2021, at 10:14, Paul Buxton via subsurface <
>>> subsurface at subsurface-divelog.org> wrote:
>>>
>>> I have been eyeing up the Paralenz Vaquita camera, mainly due to the
>>> depth based colour correction (something I had been thinking of doing
>>> myself). I emailed Paralenz about the possibility of extracting the dive
>>> log. I just got a response that it is saved out as a csv file, so we ought
>>> to be able to just use the CSV import function straight out of the box! :-)
>>>
>>> I have asked if they could send me a sample file. If I get that I will
>>> update on what it contains.
>>>
>>>
>>> thanks a lot. That is good news. Given that we have already another
>>> paralenz feature in the form of exporting depth and time as a subtitle file
>>> that can be used together with videos short during the dive for in picture
>>> display of these values, one could contemplate to also do such color
>>> correction. I am sure this would be possible with ffmepg. Maybe an
>>> interesting project for a student to research this… too bad we don’t do
>>> GSOC anymore.
>>>
>>> I myself don’t do underwater video/photo. Do you have any examples
>>> (video/pictures with accompanying depth data from a dive computer) that one
>>> could use to get started?
>>>
>>> Best
>>> Robert
>>>
>> _______________________________________________
>> subsurface mailing list
>> subsurface at subsurface-divelog.org
>> http://lists.subsurface-divelog.org/cgi-bin/mailman/listinfo/subsurface
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.subsurface-divelog.org/pipermail/subsurface/attachments/20210720/22553ffa/attachment.htm>