Mainpage
  Stereovision
  Segmentation
  Publications
  Online-Processing

Links Contact
How to Use

^ Back to Calculation-Page ^

How to use this page

General remarks
  • Fill in the appropriate data in all the form-fields of the previous page. We already filled in an example that should work.
  • After submitting your job, our server will try to fetch your images, convert them into an internal floating-point image format, do some preliminary checks and finally start the calculations.
  • You will be shown a log-file of your request, where you can watch this process. The results will simply be attached to the end of the log-file.
  • If your browser can not handle automatic updates (most browsers can), you have to push the RELOAD-button of your browser once in a while to get the current status.
  • After your calculations are finished, the results are stored at our server for further reference. However, after one or two days, the data will no longer be available to you.
Specific remarks
  • Preprocessing:
    The implemented algorithm has a specific range over which the correlation-measure is maximized. similar to human stereo vision.
    You can select either:
    • Automatic Vergence Control: a vergence system tries to align the stereo images automatically. This might fail, in which case you should try the "Fixed Disparity Range"-setting instead. The entry total disparity range sets the search region for the algorithm, which is restricted to maximal 20 pixels. In some cases you might get better results if you disable Allow for rotation, which constrains the vergence system to use only horizontal or vertical panning for alignment.
    • Fixed Disparity Range: if you know that your images have no vertical disparity component, and if you know the range of horizontal disparities in advance, this is the setting for you.

  • Correlation-Measure:
    You can choose between two basic correlation measures:
    • Sum-of-Squared Differences (SSD): as the name suggests, the squared intensity differences between left and right image patch is summed. The algorithm tries to minimize this value.
    • Correlation: - basically the scalar product between the image intensity of the left and right image patches, interpreted as vectors. The algorithm tries to maximize this value.
    Try to compare the performance of both algorithms with the algorithms where the mean image intensity is subtracted before disparity calculation. These algorithms should perform better because of the normalization of image intensities.
  • Postprocessing:
    Normally, the disparity map is only integer-valued. In order to obtain disparity estimates with sub-pixel precision, two methods can be used:
    • Subpixel-Interpolation - a simple, but effective method, except if the image-data is composed of very high spatial frequencies. The minimum of the correlation measure is calculated out of three neighbouring measurements by fitting a low-order polynomial through the data.
    • Iterative Refinement - an iterative postprocessing algorithm is applied, which tries to refine the raw estimates.

When things go wrong

  • Most problems occur in accessing your data by our server. We can handle http-, ftp- and gopher-protocol. Be sure to give a correct URL. You might check the URL you supplied to our software by opening this location in your own browser. But be sure to check by opening over the WEB, not locally. If your browser can fetch your data, the URL should be fine. If our server still can't fetch your data, please contact us.
  • The algorithms have been designed to be fairly bullet-proof, but if you do not enter correct values, you won't get any sensible results.
  • This server is used also other purposes. If a certain load-average is exceeded, this service is temporarily suspended.

© 1994-2000 - all rights reserved.