Jump to content

All Activity

This stream auto-updates     

  1. Last week
  2. cdschroer

    The size of dome

    Well, you can get decent visualizations if you violate this rule, but you won't have good normals, as you move farther away from the center of the subject. If you use normals visualization or specular enhancement with no color (both rely heavily on the normals data) you will see problems around the edges. So, it's all about "fit for purpose". If you want good normals for the whole subject, then you need the standoff. If you just want some relighting to get a decent screen shot, you might get away with less.
  3. Earlier
  4. Photosbykev

    An alternative for controlling 3W leds

    Your site gave me the inspiration to start my own little project Because my dome will only get used infrequently I've changed the fully automated system you've built to ease my build time. I'm going down the route of a dome with only seven 3W leds inside on a radial line from the bottom to the top of the dome. An arduino will control and synchronise the LEDs and triggering of a Canon 5D4 camera. The duration of the LEDs will be defined by the camera shutter by sending a signal from the camera back to the arduino so it turns the LED off when the shutter closes and the sequence moves on to the next LED. Once the first set of seven images are complete I'll manually rotate the dome 'x' degrees and shoot the next sequence. The dome is going to be mounted on a separate base so it is completely separated from the subject and the camera. Kev
  5. leszekp

    An alternative for controlling 3W leds

    You might give this site a look (not that I'm biased or anything): https://hackaday.io/project/11951-affordable-reflectance-transformation-imaging-dome
  6. Photosbykev

    An alternative for controlling 3W leds

    The Keyes 3W modules seem to be very optimistic on power rating. They only pull 0.17A at 5.7V so loosely only 1W. Using a Colorchecker calibration and setting a white balance on a grey square they are also running at 8200k. I also have a box full of Bridgelux 6500k Daylight White 3W star pcb mounted LEDs. The LED board on the KEYES device is only two solder leads onto a star pcb and very simple to remove and replace with a Bridgelux module. The Bridgelux LCD is pulling 0.4A at 5.7v so 2.3W and is approximately one stop brighter and runs with a colour temperature of 6750K. I suspect the 5.7V value is probably a bit low as I'm using an commercial LED driver to power the LEDS and the voltage waveform is likely to have some ripple on it. The dome build will continue with the Bridgelux 6500K leds
  7. Hi, probably old news for the experienced Dome users but I've just found a link to a 3W LED mounted on a pcb which only requires power and one control pin. Rather than using MOSFETs etc etc a dome could be built using only one LED driver board with all the LEDS wired in parallel and a single control wire for each LED https://www.ebay.co.uk/itm/3W-High-Power-KEYES-LED-Module-With-PCB-Chassis-For-Arduino-STM32-AVR-UK/232870808490?ssPageName=STRK%3AMEBIDX%3AIT&var=532402693460&_trksid=p2057872.m2749.l2649 I've ordered a handful to have a play with Kev
  8. Photosbykev

    Embedding Rti viewer into Wordpress

    Thank you leszekp, that's exactly what I wanted I didn't consider putting another html page into an iframe but it works nicely http://www.photosbykev.co.uk/2018/08/22/rti-imaging-baby-steps/ regards Kev
  9. leszekp

    The size of dome

    This is a useful rule-of-thumb, but I've found that you can sometimes stretch this a bit. I just did RTI on a 7" long Mayan lithic blade inside a dome with a 9" light to object distance, and got a more than acceptable photo out of it (with a bit of work in Photoshop).
  10. Katie Kremnitzer

    Error Dump

    Thank you Dave and Carla for the quick responses! Changing the path to the ptmviewer was the fix! Thanks for catching that mistake!
  11. cdschroer

    The size of dome

    Dear weiwei - do not worry. You set up a new topic, which was appropriate, and there is some information here that hopefully is useful. Also others should be able to find this, and I deleted the duplicate. Sometimes it is helpful for me to remind people, since this does happen from time to time. Welcome to the forums! Carla
  12. Hello ! I have read the"Polynomial Texture Map (.ptm) File Format" and the "RTI Format".I can realize the format from the documents.But I have a problem.I want to know where and how I could find the format? I get a .ptm or .rti file and I could view using RTIViewer. But I want to see the data content described in the "Polynomial Texture Map (.ptm) File Format".Please give me a help . Thank you very much! WeiWei
  13. weiwei

    The size of dome

    I am so sorry post the same question in more than one place on the forums.This is my first to post the question,so I will find the most appropriate forum. Thank you for your replay.It is very helpful for me .
  14. cdschroer

    The size of dome

    The recommended standoff of the lights for RTI is 2 -4 times the diagonal size of the subject. This is described in the Guide to Highlight Image Capture on page 23 (see below) If you use less than 2 times the size of the subject, you will not get accurate surface normals for the area outside the center of the subject that doesn't meet the light standoff requirement. I'm asking one of the computer graphics folks to reply as to the technical reasons for this. I know from experience and testing that this is true. Carla PS: I removed this identical question which you added to the end of another unrelated topic in the photogrammetry forum. This is appropriately a separate topic, and the discussion should take place here. As a reminder to folks - it is not helpful to post the same question in more than one place on the forums. Please find the most appropriate forum, and either add to an existing topic (if your question is part of that topic) or start a new topic. Thanks!
  15. weiwei

    The size of dome

    Hello !I have a question.I hope you will help me . I wonder know why the radius of the dome is three times the diameter of the target?Is there any theoretical basis?
  16. cdschroer

    Error Dump

    Katie - It looks like you gave RTIBuilder the path to the ptmviewer instead of the ptmfitter. You need to change it to the ptmfitter - which you also download form HPLabs. It may have been in the same package you already downloaded. In that case, just open your existing RTIBuilder file for your project (instructions towards the end of the User Guide - for "open existing project) and when you get the final screen where you can "execute" building your RTI or PTM, change the path to the ptmfitter there.'' Carla
  17. weiwei

    The size of dome

    Hello !I have a question.I hope you will help me . I wonder know why the radius of the dome is three times the diameter of the target?Is there any theoretical basis?
  18. leszekp

    Embedding Rti viewer into Wordpress

    You can embed it in a WP page using an iframe tag. Example: <iframe src="http://swvirtualmuseum.nau.edu/wp/RTI/Powell_Watch_2.html" width="600" height="640" frameborder="0" allowfullscreen="allowfullscreen"></iframe> View the result on this page: http://swvirtualmuseum.nau.edu/wp/index.php/national-parks/grand-canyon-national-park/rti-gallery/
  19. NGreene

    Embedding Rti viewer into Wordpress

    I'm also interested in this; however, I don't know a lick about coding...
  20. Photosbykev

    Embedding Rti viewer into Wordpress

    Searching the forums all of the options for viewing rti format images from within WordPress seems to have stalled years ago. I've just got the WebRtiviewer installed on my server and it works well on a static html page. Looking at the source code of the static page there are a number of links to things like spiderGL directories etc and then a simple container script to launch the rti file into the viewer. Has anyone had success porting this html code into a WordPress page? It doesn't look too difficult but I'm failing at the moment lol Regards Kev
  21. Dave Martin

    Error Dump

    Katie - have you downloaded the PTM from the Hewlett Packard research site?
  22. Katie Kremnitzer

    Error Dump

    I am not sure if this post is in the proper thread but I am having trouble executing PTMs. I was able to make HSHs but when I tried to make PTMS got the attached error message. Any idea how to clear it up? Thanks in advance!
  23. cdschroer

    Masking trouble in Agisoft Photoscan

    Thanks Dave - yes - we try to keep to more general photogrammetry issues here in this space because Agisoft runs forums for PhotoScan. We are happy to entertain general questions about setup, good practice, equipment, photogrammetric approaches, etc. That said - there are some odd behaviors within photoscan related to masks whether you are doing an individual image, or a whole set. Crystal is correct that you need to select the folder containing the images, and the background image. If that folder is already selected then I usually select another folder, then the one I want to use before saying OK. Go to the Agisoft forums if you continue to have difficulties - and make sure you include what version of photoscan you are using, and what platform you are on (Mac, windows, linux - and OS level) Carla
  24. cdschroer

    Why my rti file is invaild and internal format unknown

    Hernandez - I'm not clear what problem you are having. First you need to have a .ptm or a .rti file. If instead you have a file of some other type (like Fmck showed) then this isn't a valid rti file. My guess is this is the result of something like the WedRTIViewer which does pre-processing on an rti or ptm file to set it up to work with a web based viewer. If that is the case - then the resulting folder has to be set up on a webserver with the appropraite HTML code to call it. This is described on the web page for the WebRTIViewer. I'm not an expert on that. If you want to use the regular RTiViewer - then you need an rti or ptm file. Carla
  25. cdschroer

    Missing file: PTMFitter.exe

    Loa - We wish we could include the PTMfitter software, but this license issue (from HP) prevents it. The easiest way to check for spaces in your pathname is to go to the "assembly-files" folder created by RTIBuilder and open one of the .lp files (it is just a text file - you can open it in any text editor) There you will see the full path to the image files. You will be able to see if you have any spaces. If you do - move the whole folder (with the .xml log file, and the jpeg-exports folder, etc) to a correct location - or rename the needed directories. Then you should be able to build an RTI from the existing project without having to detect the spheres and highlights again. How to do that is described towards the end of the user guide - and also elsewhere in the forums. If the filenames themselves have spaces, then you will have to start over after renaming them. Good luck!
  26. Khalid - I can get both files from these links. The RTI spec was uploaded to this forum - so you have to be logged in to the forum to get it. I should have specified that in my earlier post. What happens if you click on the link? (For me, it downloads the file to my downloads folder) Carla
  27. hecoules

    Color mapping to normals direction

    I'm arriving a bit late to the party here, but I think that this has something to do with the problems described by Lindsay MacDonald in p106 of his PhD thesis. He claims (and I'd agree) that there's an error in the original 2001 paper by Malzbender et al. - specifically that the surface normal is stated incorrectly in eq. 18. It looks as though this error has persisted into PTM fitter, and so what is reported by RTIBuilder to be the surface normal vector is actually the specular vector. This means that if you take the 'surface normals' image from RTIBuilder and apply the linear mapping described by several contributors above (i.e. [0 to 255] for [-1 to 1] for each of the three vector components), the result is out by a factor of roughly 2 (assuming that the camera is remote from the subject). Can anyone else confirm this? Maybe someone involved with fitter development can lend some insight. Looking at the results from the HSH fitter there seems to be a similar thing going on, although in that case there is no mathematical reason (that I can think of) why there should be an issue...
  1. Load more activity
×