Forum

Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Messages - DCK

Pages: [1] 2 3 4
1
General / Re: align photos default settings
« on: March 14, 2021, 05:48:54 PM »
Thanks, Alex.

2
General / Re: align photos default settings
« on: March 14, 2021, 03:39:33 AM »
GrinGEO,

Thank you for putting so much thought into you response. There is definitely information I can use. However, my question is much, much simpler...

After importing photos, the next step in a basic workflow is to align them. When you select "Align Photos" from workflow, the option menu shown in the attached image appears. I had been playing with the values and toggles in that menu. Now I'd like it to be set back to the default. Perhaps it is. I have no idea. I would like to know the software default for this menu.


3
General / align photos default settings
« on: March 13, 2021, 12:45:36 AM »
Can someone tell me the general AND advanced default settings for the align photos portion of the workflow. I was messing around and now I'd like to go back to the default. Thank you.

4
General / Re: scale does not export
« on: December 16, 2014, 08:22:11 AM »
Nope! Where do I optimize?

5
General / scale does not export
« on: December 15, 2014, 10:31:39 PM »
The scale I set in PhotoScan does not seem to carry over when I export models to third party software (Meshlab or Geomagic).

In PhotoScan, I set scale on a model by placing two points. I know the distance between these points. I highlight the two points in the Ground Control panel, then select "Create Scale Bar." I enter the distance between the two points. Then I export the model in ply.

The scale is not retained in Geomagic or Meshlab. Is there a step I am missing?

A separate question: in the Ground Control settings, I am able to change the measurement accuracy settings. For example, the default scale bar accuracy is 0.001 m. Do I get better scale bar accuracy if I change this setting to 0.000001?

6
General / Re: ply in ascii format
« on: December 15, 2014, 10:11:57 PM »
Alexey,

Is "fix topology" only available in v 1.1? I do not see the option in the Tools menu, and i do not find any results when I search the v. 1.0 user manual for "topology".

7
General / ply in ascii format
« on: December 15, 2014, 08:04:44 AM »
I am trying to export PhotoScan models to a a program (the geomorph package in R) that requires ply files to be in ascii format. On export, I unselected the "binary" option, hoping this would do the trick.

When I attempted to open the files in geomorph, the files were unreadable. The error message says the object's dimensions are greater than its length. I spoke with one of the developers of geomorph, and she said she thinks the ply file has duplicated vertices and so it is not able to match up the vertex coordinates to the triangles

Basically, I'd just like to find a solution. Any ideas?

8
Long, narrow objects can be a little tricky. You want as many pixels per image as possible devoted to the object (as opposed to the background), and if possible, you want the whole object in focus.

One way to maximize object-pixels per image is this: Stand the object up vertically on the turntable by placing it on some plasticine clay, or in a sandbox, with the long axis of the object directly over the center of the turntable (so that the object rotates around the center of the long axis). Take pictures in "portrait" rather than "landscape" view. This fills as much of the frame as possible with your object.

Alternatively, you can shoot the object in parts, process these parts in "chunks," and then align and merge the chunks.

For the hammer, I think you can do just fine with the first approach.

The attached screenshot will give you an idea for shooting pattern. If you look closely, you can see 4 rotations around the object. The top two rotations are shot with the object resting on its base. The bottom two rotations are shot with the object resting on its distal end. So, after two rotations, I turn the object over. You should do the same. Probably the most proximal and distal rotations should be higher and lower, respectively, than what you see here, in order to get better coverage of the ends; in addition, note that these proximal and distal rotations have fewer images than the two more central rotations. That's just for efficiency. For something like your hammer you ought to be able to get away with about the same approach.

I bet three rotations is sufficient. One at the midpoint of the long axis. Two at the extremes, with coverage of the ends.

Any basic lens should be sufficient. If you're having problems with glare, for example from lacquer or because the metal is shiny, tend to your lighting or consider buying a circular polarizing filter for your lens. They're cheap.

9
General / Re: GPU fail changing to CPU
« on: October 03, 2014, 08:49:12 PM »
Quote
And regarding the second question, it is recommended to enable all CPU cores in case there are no OpenCL devices enabled.

Alexey,

I thought it was recommended to disable one CPU core per GPU enabled, and double this amount if you have hyper-threading. So, with one GPU, 8 CPU cores, and hyperthreading, one should have 6 CPU cores active. Am I mistaken?

10
FYI I use PS Professional edition, v. 1.0.4....

SETTING SCALE: I took a bunch of images with scale bars affixed to the turntable. I'd like to use these scale bars to set scale on the object itself.

Questions:
1. Do I place the markers on the scale bar in the photos themselves?
2. If not, I guess that means I need to model the scale bar as well as the object, and then place the markers on the scale bar, yes?
3. If so, in how many photos do I need to model the scale bar?
4. If I have more than one chunk, do I need a separate scale bar from each chunk?
5. Alternatively, I could place markers on the model of the object, measure the distance between those locations on the object itself, then set scale by scaling the object to that distance between markers. In fact, this is the approach I have currently taken. Is this as accurate as using the scale bar? Let's say I take three measurements on the object itself. For simplicity, let's say I take a length, width, and height measurement. Each measurement has error. If I place markers on the model to replicate these three distances, because of measurement error, each distance will suggest a slightly different scale for the object. How does PS handle these conflicting scale measurements? Averaging?

SURFACE AREA AND VOLUME: Having set scale for my object using a single measurement, in the manner described in question 5 above, I would now like to estimate surface area and volume for my object. To do so, the PS Manual instructs me to choose the "Measure Area and Volume" from the Tools menu. My tools menu does not have a Measure Area and Volume option. Where is it?

Screenshot attached. It's a metatarsal (toe bone)... get your mind out of the gutter :)

Thanks.

11
I doubt you'd need the newspaper to photograph the hammer. It appears to be distinct enough.

Square Perfect makes good, cheap light tents. Take a moment to find an online video about how to fold them up when you're finished. It's a bit non-obvious.

Virtually any TV turntable will do. Get one with a matted finish however. If you want to use newspaper for alignment, tape the newspaper to the turntable.

12
General / Re: how to creat Scale Bar
« on: October 03, 2014, 01:46:27 AM »
That helps somewhat. But it suggests that I need to actually model my scale bar. I would prefer not to.

1. Is this possible?
2. If I have multiple chunks, do I need to set scale for each?

13
General / Re: how to creat Scale Bar
« on: October 03, 2014, 12:06:51 AM »
I'm having trouble following the scale bar instructions in the user manual. This thread contains a link to a tutorial on setting scale. However, that link is broken. Is it possible to re-establish it please?... or a link to any thread or reference that walks step by step through setting up scale for a model.

Thanks.

14
Bug Reports / Post attachments bug
« on: September 30, 2014, 04:44:37 AM »
There's a bug in bug reporting on your new website. The post will not accept attachments.

15
General / Re: Using physical markers in "background" to improve alignment
« on: September 28, 2014, 02:16:25 AM »
Here are the not-so-stellar results. In the solid and shaded views, you can see the skull, the plasticine, and the black wooden surface. In reality, the skull is by far the smoothest of these three surfaces; in the model, the skull is bumpiest.

Camera alignment looks great, however. :/

I think it is simply that the smooth white surface is not easily modeled. This is obvious at every stage from sparse cloud to shaded. I can improve results slightly by gradual selection on the sparse cloud, but not to the level where the model actually resembles the object.

Results are similar with Nikon 12mpx and Canon 21mpx.

Thanks to everyone who's tried to help. Always open to additional suggestions.

Pages: [1] 2 3 4