Author Topic: How to clean up the SRC 'Deadzone?  (Read 275 times)

0 Members and 1 Guest are viewing this topic.

Offline Fishton

  • FishTec
  • Jr. Member
  • **
  • Joined: Mar 2010
  • Location: Westville
  • Posts: 53
    • FishTec - Sonar and Mapping Solutions
  • Unit(s): 1197cxSI
How to clean up the SRC 'Deadzone?
« on: March 03, 2019, 10:28:47 AM »
I have been trying to get a 'fix' for the dead zone where the water column has been removed by the Slant Range Correction.

We know that the sonar beam is not just a 90º side to 90º down, so why can't each side collect just a few feet more, where left overlaps slightly into right and right slightly into left?

This will allow the SRC to clean up that fuzzy dead zone where the left and right mosaic meet, thus removing that unsightly 'trail'.

To remove this with a photo editor and only use a single side to 'tile' the mosaic is just too time consuming, especially on large bodies of water.

Have been asking Tore (support@sonartrx.com) for a few years now for any help regarding this matter but to no avail.

Has anybody got any ideas regarding this?

Regards
John


Offline Rickard

  • Hero Member
  • *****
  • Joined: May 2009
  • Location: Mariefred, Sweden
  • Posts: 508
  • Unit(s): 999, 981, M37, LowBird-1, LowBird-2
Re: How to clean up the SRC 'Deadzone?
« Reply #1 on: March 04, 2019, 03:54:40 PM »
Hi John,

As far as I know there is no reliable way to clean up SI data from the sector right below the transducer. Firstly, a sonar image is primarily composed of shadows, but in that direction it's almost impossible to get any shadows. Secondly, all positions on the surface in that sector is located at nearly the same distance from the transducer, so the raw image is extremely compressed. Thirdly, since a SI channel collects some data from the opposite SI channel's area, using it's own transmitted signal, those data are mixed with data from "it's own" side. The transducer doesn't know from which direction echoes arrive, it just records time and intensity. There is no way that I'm aware of to separate data afterwards. A photo editor doesn't help, because the image can't show anything to cut away because the mixed data are displayed at the same place. Slant range corrected images will always be blurred below the boat.

Regards,
Rickard

Offline Fishton

  • FishTec
  • Jr. Member
  • **
  • Joined: Mar 2010
  • Location: Westville
  • Posts: 53
    • FishTec - Sonar and Mapping Solutions
  • Unit(s): 1197cxSI
Re: How to clean up the SRC 'Deadzone?
« Reply #2 on: March 04, 2019, 08:56:11 PM »
Hi Rickard,

What you said makes total sense, I must confess, I forgot all about the shadow.

My intension now is to place the outer pontoons on my trimaran mapping craft around 10ft apart, with the left scanning transducer on the right and the right scanning transducer on the left.

As my mapping is shallow water only, I believe this will create just enough of a shadow to cleanup the SRC.

Then surely just a small software tweak will put it all together quite nicely?

Regards
John


Offline Rickard

  • Hero Member
  • *****
  • Joined: May 2009
  • Location: Mariefred, Sweden
  • Posts: 508
  • Unit(s): 999, 981, M37, LowBird-1, LowBird-2
Re: How to clean up the SRC 'Deadzone?
« Reply #3 on: March 05, 2019, 02:32:55 AM »
Yes, this will give better shadows from the area below the middle of your trimaran in each channel. But the channels will also interfere with each other, if they use the same frequency. To avoid this you could locate one channel at the bow and the other at the stern. But then the channels will recode data from a specific position at different points in time, so that time lag must be compensated to make it possible to join the left and right images. After slant range correction the worst part of the "dead zone" can be removed. Just cut away a strip equally wide as half the distance between the left and right transducer from each image. But, the critical point in this process is compensation of time lag. I think this is a hard thing to solve for the software people. Speed and position data must be very precise to make it meaningful.

/Rickard

Offline Fishton

  • FishTec
  • Jr. Member
  • **
  • Joined: Mar 2010
  • Location: Westville
  • Posts: 53
    • FishTec - Sonar and Mapping Solutions
  • Unit(s): 1197cxSI
Re: How to clean up the SRC 'Deadzone?
« Reply #4 on: March 05, 2019, 03:33:06 AM »
I record off a single unit with the GPS puck dead centre and above the P66 200kHz 9º cone on the centre hull stern.

Then for side I use two transducers but wired to the same single plug the P66 uses, so the unit still thinks it is using a single transducer.

This will not be difficult to correct on SonarTRX, all it needs is to process the left and right separately but allowing a 'Crop' at the beginning of the mosaic, not just the end as it currently has.

Offline Rickard

  • Hero Member
  • *****
  • Joined: May 2009
  • Location: Mariefred, Sweden
  • Posts: 508
  • Unit(s): 999, 981, M37, LowBird-1, LowBird-2
Re: How to clean up the SRC 'Deadzone?
« Reply #5 on: March 05, 2019, 05:03:47 AM »
But have you tested this with a "manual" procedure? This would tell if influence from interference between the SI transducers is a problem. If it's not, then a program that only needs an extra manual setting of inter-transducer distance seems possible. Of course, a program developer must consider the market, is it big enough to make it worth the effort?

I looked at some of my slant range corrected recordings, they show bad zone up to about 1/3 of depth. This means the distance between SI transducers must be at least 2 * 1/3 = 2/3 of depth. (The right channel transducer must be moved 1/3 of depth to left to deliver good recording from the middle of the boat and beyond. The left channel transducer must be moved the same distance in the opposite direction. The sum of these distances will be 2/3 of depth.) With a 10 ft wide vessel this means one can map up to 15 ft depth with good results. At large depth, this method becomes problematic. At 60 ft (20 m) the transducers must be located 40 ft (13 m) apart.

/Rickard

Offline Fishton

  • FishTec
  • Jr. Member
  • **
  • Joined: Mar 2010
  • Location: Westville
  • Posts: 53
    • FishTec - Sonar and Mapping Solutions
  • Unit(s): 1197cxSI
Re: How to clean up the SRC 'Deadzone?
« Reply #6 on: March 05, 2019, 05:28:45 AM »
Mmmmmmm, I see the problem here.

Multi 'tiling' type passes seems to be the solution, but double the work as you can only do one left/right, then just one side after that.

Autonomous mapping would make the passes more accurate, but GPS lately is very erratic when it comes to EPE.

Offline Fishton

  • FishTec
  • Jr. Member
  • **
  • Joined: Mar 2010
  • Location: Westville
  • Posts: 53
    • FishTec - Sonar and Mapping Solutions
  • Unit(s): 1197cxSI
Re: How to clean up the SRC 'Deadzone?
« Reply #7 on: March 05, 2019, 06:20:57 AM »
But I still see no reason why the Crop tool under 'Create Mosaic' cannot include a 'beginning' of mosaic as well as end of mosaic crop.  ReefMaster has it.


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo

 

Related Topics

  Subject / Started by Replies Last post
29 Replies
18211 Views
Last post May 30, 2010, 03:04:40 PM
by vmoyers
6 Replies
7168 Views
Last post May 30, 2010, 02:57:38 PM
by vmoyers
0 Replies
6831 Views
Last post May 31, 2012, 04:48:14 PM
by sonar2000
5 Replies
3443 Views
Last post May 06, 2013, 03:56:01 PM
by Humminbird_Greg
3 Replies
3573 Views
Last post December 16, 2013, 04:24:40 AM
by Rüdiger


SimplePortal 2.3.3 © 2008-2010, SimplePortal