Changes between Version 8 and Version 9 of Procedures/AlsprocProcessing

Sep 26, 2013 2:47:07 PM (5 years ago)



  • Procedures/AlsprocProcessing

    v8 v9  
     137Then go into the dataset and look for rooftops or other features both along and across areas of overlap. It should look like this:
     139[[Image(pitch_error_249-2012.jpg, align=center)]]
     141Using the ruler tool come up with the best possible approximate for lateral displacement both along and across the overlap. Normally this will be in the region of 50cm to 10cm. If the displacement is less than 5cm or 10cm in both directions then the files you have should be ok to use and you can use the pitch and roll errors you used for this attempt in your final solution. If not, you will need to estimate some new pitch and roll errors and try again. An example of acceptable pitch and roll errors is given below.
     143[[Image(pitch_error_corrected_249-2012.jpg, align=center)]]
     145Determine the direction of your error. You need to decide whether to add or subtract from each of the pitch and roll error. Look at the times of a few points in each flightline to determine each of their directions, and then decide whether you want to roll the angle of the laser clockwise or anticlockwise around the direction of travel of the plane to move the data left or right respectively; and whether you want to pitch up or down to move the data forwards or backwards respectively.
     147The table below indicates the effects of addition of subtraction on each axis.
     149|| '''Operation''' || '''Pitch''' || '''Roll''' ||
     150|| addition        || ? || ? ||
     151|| subtraction     || ? || ? ||
     153You may estimate the magnitude of your adjustment yourself based on your own intuition or previous observations if you wish. Using an equation can help you zero in on the correct solution more quickly, especially in your first two or so observations.
     155e = arctan( d / (h1 + h2) )
     157Where: e will be your error to add to or subtract from the value used to compute this dataset, d is the difference you measured in lag, h1 and h2 are the heights of the plane in each flightline above the terrain at the time of the observation.
     159Now return to processing the dataset, you can inspect the data and pass it off, or you can perform another iteration to get the match even closer. The number of iterations you will need to do can vary, but often depends on the quality of the navigation data.