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 - stihl

Pages: 1 ... 4 5 [6] 7 8 ... 28
76
General / Re: Problem with overlapping images !
« on: May 16, 2017, 04:26:55 PM »
Hi Ramadnio,

What you'd like to do is create a Mesh based on the *ground* points of the Dense cloud.

This will ensure that you can create an artifact free Orthomosaic.

After generating the Dense Point Cloud, run the automated classify ground points alghoritm with the default parameters.

Then generate a new Mesh with only these ground points. This can be done in the Build Mesh window.

Once that's done simply doubleclick on the Orthomosaic in the workspace pane and create polygons around objects and buildings to manually select an appropriate image for that particular polygon.

77
General / Re: DSM Generation
« on: May 10, 2017, 01:09:52 AM »
DSM = DEM = height model of everything

DTM = height model of merely the terrain

78
The spacebar on your keyboard acts as a hotkey, letting you toggle between navigation mode and polygon/polyline/single point mode. Any of the three will switch back to navigation mode once the spacebar is pressed.

79
Hi Shovelhead,

Surveyor here for 7 years and using photoscan for 5 years.

From experience I can tell you that the absolute accuracy you can achieve with photoscan is higher than pix4D.
This is because photoscan allows more user input such as filtering of tie-poinrs and more extensive model build parameters.

The things you've seen in the link you posted can also be done in Photoscan by simply using the polyline or polygon tool on the toolbar in the photoscan interface. You can directly draw lines in the point cloud or in the mesh model and export these directly to a shapefile or kml/kmz.

80
General / Re: Uneven performance (samples/sec)
« on: April 26, 2017, 12:55:59 PM »
My rough guess would be that it has to do with the temperature of the laptop laptop. I can easily imagine that it heats up quite a bit as laptop usually tend to have poor airflow to begin with. Combined with a relatively high clocked CPU perhaps the laptop tunes down the CPU speed to increase stability over time.

81
Hi,

I don't think there's anything wrong. Photoscan uses the RAM of your computer to load the necessary data quickly for processing.
A project of a thousand images (24mpx) may not need more memory consumption than the values you've mentioned. There's a benefit to having some overhead of RAM for the computer while processing.

I assume your CPU is nearly always at full load during processing?

82
General / Re: DEM of Top of Canopy?
« on: March 26, 2017, 03:18:04 AM »
Hi!

As far as I know those kind of analyses aren't yet available in photoscan pro.

I'd advice to export the DEM as a TIFF and load it into GIS software such as Global mapper/QGIS/ArcGIS.

Determining how to do the specific analyses in those sofware packages might be better answered on their respected forums.

83
General / Re: Duration of photo alignment
« on: March 23, 2017, 01:39:08 AM »
Using generic preselection will greatly reduce the processing time needed. I'd only advice not to use it when generic mode fails to align the dataset properly.

84
General / Re: View project history while processing in progress?
« on: March 10, 2017, 12:23:44 AM »
In Pre 1.3 versions I had no problems opening up a second instance of the same running project.
So if you saved  your project between alignment and pressing on generate dense cloud, you can simply open the project again and see what you did in the Chunk Info window.

85
General / Re: GSD and GCP in very steep terrains
« on: March 07, 2017, 04:19:10 PM »
If you can guarantee the points that you mentioned then I'm certain you will yield better results with scenario A.

Out of curiosity; which drone will you be using?

Good luck with your project!

86
General / Re: GSD and GCP in very steep terrains
« on: March 07, 2017, 03:33:20 PM »
Hi,

In scenario A you mentioned that the drone can follow the terrain altitudes. However where does it get this elevation information from and how updated is this data? This is quite crucial for performing the flight according to scenario A. If this terrain information is inaccurate this will cause the drone to experience minimal or insufficient photo overlap for reconstructing high elevation areas because the programmed amount of photo overlap will be reduced when you get closer to your terrain.

If this terrain information is accurate then you will likely get the most uniform results using this scenario.
Also I believe you mean that the drone altitude is relative to the WGS84 ellipsoid which is different from relative mean sea level that you mentioned.


On a side note; empirically I've found out that the amount of GCPs needed depends more on the shape and dimensions of the project area than the value of the GSD. For instance long narrow project areas require more GCPs than a project area that's basically a square.

87
General / Re: Report DEM Resolution
« on: February 08, 2017, 07:50:19 PM »
A DEM is an IDW interpolated model of the dense point cloud. If the resolution of the DEM is 10 centimeters, this means it's made up out of 10*10cm box grid. Meaning that the resolution is the same throughout, also in a vertical plane.

88
General / Re: Scaling a mesh using a 0.5m scale bar
« on: February 02, 2017, 01:31:27 PM »
Hi Tom,

It's quite easy. First double click on a photo in your photos pane that has the scalebar in it. Then zoom in on one side of the scalebar and right click at the place where the scalebar starts and select Create Marker.
Then click on another photo in the photos pane that has that same scalebar in view and zoom in on the same place on it. Now right click and select Place Marker and select Point 1 (default name).
To the left in the markers panel, this Point 1 has now appeared. Right click on it and select Filter Photos by Markers.

The photos pane will now only show photos where Point 1 is visible. Mark all the photos that has Point 1 in it.
Now do the exact same thing for the other side of the scalebar with a newly created marker and mark the other side of the scalebar in as many of photos as well.

Once that's done you select both markers in the marker pane, right click and select Create Scalebar.
This will create a virtual scalebar which will show up in the Scale Bars pane. Double click on the empty value below Distance (m) and fill in the exact length of the scalebar you used in your project. After that optimize your project with the Optimize Cameras button in the reference pane.
If you have more scalebars, do the same thing for those as well. It's best to place two scalebars perpendicular to eachother so you have a scale for the X and the Y axis.

89
General / Re: How to find out the pixel size in Camera Calibration?
« on: February 02, 2017, 11:05:01 AM »
Taken from this website: http://www.digicamdb.com/specs/canon_ixus-125-hs/

The pixel pitch( pixel size) of your camera is 1.33 microns which is 0.0013 mm for photoscan.
Not sure where it went wrong with your calculation. But the EXIF data is usually correct. In some scenarios this information night be missing, best option then  is to look up the camera specs online.

90
General / Re: Best way to process 4500 photos?
« on: January 30, 2017, 02:08:08 PM »
With that much RAM, why would you need more than one chunk?

You don't. Although I appreciate the efforts by a1mless above, I think it's best if he were to align it in one chunk.

Before aligning make sure that the camera calibration parameters are filled in (such as focal length and pixel size) and then align with High accuracy, reference pre-selection and 40k key-point and 0 tie-point limits.

Pages: 1 ... 4 5 [6] 7 8 ... 28