Localization Issue? or Page Issue

Randy Hepworth

New Member
It happened last week. I had a project setup with a Local Grid on its own Page using WGS84(ITRF2008) as the underlying coord system with imported local points. Started the project by locating several common points in SPC. I have utilized this procedure many, many times. I then Localized to three points, and all was well, I then paired two incorrect point pairs. (the GPS shot was not for the right Local Grid point number). Obviously at that point my localization went wrong. So I attempted to correct the point pairing. Not sure what happened, but at that time the selection for the Pages that were turned on went screwy. I could turn on all of the Pages, except the Local Grid page would not turn on. If I purposely chose the Local Grid page, all of the other Pages turned off!. I deleted all of the Local Grid points, imported them into a new Local Grid page, and attempted to Localize, but the Survey page would not turn on at the same time as the Local Grid page. I did a work-around that day by importing the local grid points into a SPC based page, and then translated/rotated to the GPS shots.

Now, I am attempting to set up another project today for tomorrow. Same issue: I cannot have the Local Grid page turned on at the same time as any of the other Pages! What's wrong?
 

John Thompson

Well-Known Member
Something similar happened to me this morning. I was on the PreRelease version. I went back to Release and it went away.
 

Randy Hepworth

New Member
Thanks, John- I was on the PreRelease version. Just updated the Release version and went back to all those jobs with the custom Coord System; now all those Pages will turn on with the SPC coord-based Pages. Apparently, something is wrong with the PreRelease Version with using WGS84(ITRF2008) as the underlying coord system and having that Page turned On along with all of the other Pages. Your suggestion worked.
 

Shawn Billings

Shawn Billings
5PLS
It's a new "feature". Until a coordinate system is localized you cannot have a page with the local system on at the same time as a page with a known system on. This makes selecting points from the map for localization very difficult.
 

Randy Hepworth

New Member
OK, but how can you Localize without having both pages on at the same time? That was my problem. The first time it happened (last week) I had actually started the localization process, and was refining it with additional points until I goofed and mis-matched two pairs of points. At that point I lost having both pages on. Therefore, my localization was stalled and could not continue with it. Had to do the work-around: translate/rotate the local grid within a Page set up for SPC.
 

Vladimir Prasolov

Well-Known Member
JAVAD GNSS
Sorry for inconvenience. We have to have one pair to make localization linked to some geographic area. It is important due to some digital models are working in particular geographic area (e.g. geoids). So if we will allow using undefined localization system as defined we will have calculation issues. To prevent this we could do that automatically by using current location as option.
 

John Thompson

Well-Known Member
we could do that automatically by using current location as option.
I'm not sure what you mean by this. I don't want to use the current (office) location to localize, and I don't want to wait until I've found a corner to use that location, either. I want to be able to stake the linework to find the first point. For local systems that are not yet defined by a pair of points, is it not possible to treat them as though they were the underlying coordinate system? That's how it used to work.

I started a new job today at a place I've never been before so I have no surveyed points. I have 5 polylines representing prior deeds/surveys that I have aligned with georeferenced photos, so I think I have the corners within a few feet. I imported each polyline on its own page with its own local coordinate system. I also imported a series of points that represent the initial guess of the first vertex of each polyline. These were imported on a separate page with the project coordinate system (in this case an LDP, but could be SPC...just not a local CS).

So in CAD, I have a polyline on layer DEED. I draw a point at the beginning of the polyline and name it TempDEED. I import the polyline into J-Field on Page 1 with a local CS. I import TempDEED on Page 0 with the project CS. In Localize, I select the local CS for DEED, and link DEED_1_1 with TempDEED. Now the local CS is defined. As I collect points I can link them and eventually delete TempDEED.

That way I can stake the linework to search for the corners before I have found any points. Shawn can probably think of a simpler way to do this. It's a lot more steps than before (which was already a lot of steps), but it works. I'm really hoping this new "feature" is not permanent. Localization is a very powerful part of J-Field. I use it a lot but this has made it a real pain. It's frustrating when software that used to work well doesn't work anymore.
 

Vladimir Prasolov

Well-Known Member
JAVAD GNSS
I'm not sure what you mean by this. I don't want to use the current (office) location to localize, and I don't want to wait until I've found a corner to use that location, either. I want to be able to stake the linework to find the first point. For local systems that are not yet defined by a pair of points, is it not possible to treat them as though they were the underlying coordinate system? That's how it used to work.

I started a new job today at a place I've never been before so I have no surveyed points. I have 5 polylines representing prior deeds/surveys that I have aligned with georeferenced photos, so I think I have the corners within a few feet. I imported each polyline on its own page with its own local coordinate system. I also imported a series of points that represent the initial guess of the first vertex of each polyline. These were imported on a separate page with the project coordinate system (in this case an LDP, but could be SPC...just not a local CS).

So in CAD, I have a polyline on layer DEED. I draw a point at the beginning of the polyline and name it TempDEED. I import the polyline into J-Field on Page 1 with a local CS. I import TempDEED on Page 0 with the project CS. In Localize, I select the local CS for DEED, and link DEED_1_1 with TempDEED. Now the local CS is defined. As I collect points I can link them and eventually delete TempDEED.

That way I can stake the linework to search for the corners before I have found any points. Shawn can probably think of a simpler way to do this. It's a lot more steps than before (which was already a lot of steps), but it works. I'm really hoping this new "feature" is not permanent. Localization is a very powerful part of J-Field. I use it a lot but this has made it a real pain. It's frustrating when software that used to work well doesn't work anymore.
Dear John,

I would like to suggest following solution:

1. If underlying coordinate system is geodetic, we keep behavior of program as is. It is due to impossibility to know where is center of constructed local system.

2. If underlying coordinate system is grid, program will consider constructed local system as just tuning of underlying one. Program will give notice about possible issues with digital models in case of substantial offsets. Optionally program will allow input offsets.

Please let us know if this is acceptable solution.
 

Randy Hepworth

New Member
Back in the day- before GPS! we used to have to use top-mounted EDM on a theodolite, remember those? Actually, my field work experience began with transit & tape which transitioned to Wild T-2 & HP-3800 EDM. But I have a huge legacy of local coordinate file projects going back to 1988 when computers began for me. I then eventually graduated to a base/rover Trimble R-8 system. The Trimble system allowed the user to to "calibrate" to a local coordinate base (5000,5000) project which was very similar in operation with JAVAD localization. I have a library of 14 years worth of SPC coordinates collected from hundreds of projects and have been doing Calibration/Localization for 14 years. It should be a simple matter of pairing common points of "local grid" with GPS collected SPC. After getting the right procedure from Shawn Billings after I purchased the Javad LS in 2017, I have successfully used the JAVAD procedure to localize- until an early-June update of the LS system, that is. Due to such a library of coordinates, usually I can set up the localization even before leaving the office for the project which is a huge time saver, and more importantly a money maker. By the way, I will not go back to the Trimble R-8 system; the LS works circles around it!

At any rate, I had been using the Pre-Release version when the issue appeared. After switching to the Released version with John's suggestion, I have had no more issues- Localization is working as it has for the last few years.
 

John Thompson

Well-Known Member
Dear John,

I would like to suggest following solution:

1. If underlying coordinate system is geodetic, we keep behavior of program as is. It is due to impossibility to know where is center of constructed local system.

2. If underlying coordinate system is grid, program will consider constructed local system as just tuning of underlying one. Program will give notice about possible issues with digital models in case of substantial offsets. Optionally program will allow input offsets.

Please let us know if this is acceptable solution.
That sounds like it would work. Thanks.
 
Top