0

I have got this problem, so there is a set of data as points in the spherical coordinate system - local (not a faithful arrangement of geographic or mathematical)and I'm trying to convert it to a Cartesian system to preview it in any program to draw the shape which should rise from these points .

Points are collected by the meter with a rotating laser head (thus slightly noisy). The head rotates in two axes, called phi, theta and the distance r.

Where

  • phi - is left-right rotation (-90 to 90)

  • theta - is up-down rotation (-90 to 90)

  • r - the distance

This can be seen in the figure below: img

I tried to convert the data to Cartesian (xyz) according to the following formulas: img

Unfortunately, every time something happens to me run them down and the picture that I get is incorrect. For sample collection:

sample

I get such a picture (seen from top): img

The expected result should be a rectangular tub (with bare upper part). This first arc (at the point where data has not yet been ran over) is called. lens effect, resulting from the fact that the meter was close to the wall, and second end of graph puzzles me end where the data are arranged in a completely unexpected.

With this number of points is hard for me to figure out what causes failure or bad for the conversion of data, or simply meter so measured. I would be grateful for verification my way of thinking and any advices.

Thank you in advance.

Spektre
  • 49,595
  • 11
  • 110
  • 380
pzima91
  • 1
  • 3
  • is the LIDAR/SONAR measuring from the center of rotations or it is displaced a bit (in which way direction axis only or also radially)? What are the texts in the sketch: I recognize 1. `z boku = from side` 2. `obne?motor = booth motors?` 3. `z geny = from up` as they are not english you should clarify a bit (looks like some Slavic language do not recognize it thou but I am no linguist ...). Equations looks OK so either your LIDAR/SONAR is displaced and you need to add constant or something to `r` (the curvature on the right) or something mess up your measurement. What is the sensing method? – Spektre Jul 31 '15 at 07:14
  • 1
    There is an extra factor in the y equation. How did you code that ? –  Jul 31 '15 at 09:27
  • @pzima91 the file you posted is already converted to cartesian, you should share the original spherical data. when rotating your data the bottom of the tub is also a bit funny curved – Spektre Jul 31 '15 at 15:04

1 Answers1

0

I think i am late to answer this question. I can't see the images, anyway you can go through enter link description here . It will give you clear idea how to convert spherical cordinate data into cartesian cordinate system.

Shabir
  • 107
  • 1
  • 7