Results 1 to 8 of 8

Thread: My New Structured Light Scanning Rig and first scan

  1. #1
    Join Date
    Nov 2008
    Posts
    665

    Default My New Structured Light Scanning Rig and first scan

    As soon my mini 3d led projector (Native resolution 1024x600) arrived i started making a simple scanning rig for my structured light scanning.The test piece was a complicated cherub model. Total 10 scans were manually aligned and fused. Each scan took about 8sec. The quality and alignment of the scan is impressive. Currently, the Scanning rig is under modification to get variable size scans.
    i have managed to scan this masterpiece with Vertical Negative corner and the projector-camera side by side. The results are good I have not tweaked any settings except i changed the exposure of camera 1/60sec which gives good results.Amazingly i was able to Fuse the scan with 5000 poison.Previously i was unable to pass 1000.
    As this was a first test SL scan so precision was not important for me, instead, all the 10 scans aligned one into other perfectly.
    Attached Images Attached Images
    I left woodwork due to office work.

  2. #2
    Join Date
    Mar 2008
    Location
    Tulsa Oklahoma
    Posts
    1,237

    Default

    My results with structured light are much like yours. Fast to take the pictures, but processing takes a minute or so after that.

    Resolution is limited by the larger pixel size on the target of either the camera or the projector. When I was testing my version of SL I was shooting flat subjects and looking at the variance in Z to measure errors. I dont remember the exact numbers, but the results were impressive at the time.

    At the moment I am back to having no projector.. so I have to leave that project on hold a while.

    D
    "The best thing about building something new is either you succeed or learn something. Its a win-win situation."

    --Greg Westbrook

  3. #3
    Join Date
    Sep 2008
    Location
    Newberry, MI
    Posts
    566

    Default

    Very cool Khalid! Haven't talked to you in a while! Glad to see you are still experimenting...

    Mike

  4. #4
    Join Date
    Nov 2008
    Posts
    665

    Default

    Quote Originally Posted by dana_swift View Post
    My results with structured light are much like yours. Fast to take the pictures, but processing takes a minute or so after that.

    Resolution is limited by the larger pixel size on the target of either the camera or the projector. When I was testing my version of SL I was shooting flat subjects and looking at the variance in Z to measure errors. I dont remember the exact numbers, but the results were impressive at the time.

    At the moment I am back to having no projector.. so I have to leave that project on hold a while.

    D
    Hi Dana,
    I have seen some screenshots of your SL scan you posted on this forum. I think that was a duck scan and you have shown grey height maps only, you never shared a full resolution mesh of your scan which i eager to watch
    I salute you for the programming on SL scan which seems very tough job. I really want to see some of your latest work on SL.
    Thanks for posting
    I left woodwork due to office work.

  5. #5
    Join Date
    Nov 2008
    Posts
    665

    Default

    Quote Originally Posted by mikeacg View Post
    Very cool Khalid! Haven't talked to you in a while! Glad to see you are still experimenting...

    Mike
    Thank you Mike... Office work is too much now adays.. Yesterday we have 24 hours work on maintenance/machining of control valve without any sleep.
    Today i have completed my scanning rig for David Laser scanner and did scan of grape leaf.
    Attached Images Attached Images
    I left woodwork due to office work.

  6. #6
    Join Date
    May 2008
    Location
    Port Orange Florida
    Posts
    194

    Default

    Have you tried the new David 3.5 and did it work better than 3.
    Thanks
    Don Clifton
    4axiscnc.com

  7. #7
    Join Date
    Nov 2008
    Posts
    665

    Default

    Quote Originally Posted by donclifton View Post
    Have you tried the new David 3.5 and did it work better than 3.
    Thanks
    Don Clifton
    4axiscnc.com
    the above scans are from 3.5.1. The new david is excellent in quality and precesion
    Attached Images Attached Images
    Last edited by khalid; 10-21-2012 at 01:01 PM.
    I left woodwork due to office work.

  8. #8
    Join Date
    Mar 2008
    Location
    Tulsa Oklahoma
    Posts
    1,237

    Default

    Khalid- If you recollect, I inquired about the camera you were using on the David scanner. Sometime after that, I bought a USB camera with 2500 pixels horizontal. Getting a hi-res camera was a good and a bad thing. The extra amount of pixels to process takes longer and takes more memory to store the frames. The good part is the extra resolution of course. Now my limit to resolution is the size of the projector pixels on the object.

    The current state of the project is: I investigated algorithms to determine the camera location relative to the projector. I came up with an iterative algorithm that works, but decided to investigate closed form solutions. I was successful at finding a closed form solution, but its not as simple to express as the iterative method. That ended up being an interesting diversion, but not fruitful. The process of determining the camera position from the image is called "Churches problem", as it was first solved using a similar iterative process for aerial photography to determine the location of the airplane. It was developed by a guy named Earl Church. From what I can determine Churches method is still used in that industry.

    The last thing I did before dismantling my projector and camera setup was to check the accuracy of the digitized results. I did this by scanning blocks of known size sitting on a flat surface. And in some cases against a flat wall. The flatness of the surfaces was measured to get statistics on the quality, as well as the absolute values of the errors.

    My scans of test objects like the duck were quite successful, as well as sending surfaces into Aspire to cut 1:1 copies. I don't have any photos of the results at the moment, but I'm certain I will do some more 3D work in the future. Who knows when..

    A project like this is very trig intensive, its a great solution to my needs, Now I need a commercial reason to purchase a projector with horizontal resolution around 2500 pixels so I can have a matched optics system. Then I will have to do all the accuracy measurements again.. if I remember to save the data I will pass it along. A commercial application would justify setting it back up again.

    D
    "The best thing about building something new is either you succeed or learn something. Its a win-win situation."

    --Greg Westbrook

Similar Threads

  1. SB3 Structured Tutorials
    By gerryv in forum ShopBotter Message Board
    Replies: 6
    Last Post: 08-28-2012, 03:05 PM
  2. Last Day and Laser Scan of Coin
    By khalid in forum Folder 2012
    Replies: 6
    Last Post: 04-21-2012, 11:41 AM
  3. Kinect to scan an object
    By myxpykalix in forum ShopBotter Message Board
    Replies: 7
    Last Post: 03-26-2012, 12:27 PM
  4. Looking for someone to scan some carvings
    By r_jones in forum ShopBotter Message Board
    Replies: 2
    Last Post: 09-08-2011, 12:29 PM
  5. Structured Light 3D Scanning
    By myxpykalix in forum ShopBotter Message Board
    Replies: 4
    Last Post: 01-05-2010, 01:47 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •