Gocator Interface / Driver

General Questions
1

From LMI I heard that there is some interface for the LMI Gocator series. Especially for reading out the metadata from the blue channel in RGB16 mode or de last line in the monochrome mode. Do we have any functionality to correct pixels based on the metadata and perform for instance thresholds on “height” e.g. giving all objects (blobs) above 15.2 mm?

We startet such a library, but if this is available from CVB Development Team that would be better.

2

Hi Dietmar,

there is a C# example available that displays the measured rangemap and uses the stamp data to automatically convert it into a pointcloud. However, it is still based on the old CVB 3D tools including the old display and old core3D functionality. After releasing the new 3D tools I will update the Gocator example. In the new CVB 3D functionality there will also be other 3D functions, such as a 3Dcrop with which you can crop points from a point cloud using given limits.

You can download the old example here: 134197-LMIGocatorGenTLExample_CVB2016.zip (254.9 KB)

What you need to do is to convert the stamp data into metric values first:

 // FUNCTION: Convert stamp data into metric values (each four 16bit stamp data pixel into one 64 bit stamp data value)
    private Int64[] ConvertStampData(Int16[] data)
    {
      Int64[] stampData = new Int64[15];
      Array.Clear(stampData, 0, stampData.Length);
      Int64 val1 = 0, val2 = 0, val3 = 0, val4 = 0;
      for (int i = 0; i < 60; i += 4)
      {
        val1 = data[i];
        val2 = data[i + 1];
        val3 = data[i + 2];
        val4 = data[i + 3];

        stampData[i / 4] = ((val1 & 0xFFFF) << 48) | ((val2 & 0xFFFF) << 32) | ((val3 & 0xFFFF) << 16) | (val4 & 0xFFFF);

        // Computation for x-offset correction (Gocator encodes intensities with int16, but values in .bmp image are stored in uint16)
        if (i == 44)
          stampData[(i / 4) - 1] = stampData[(i / 4) - 1] + 32768 * stampData[i / 4];
      }
      return stampData;
    }

Then you can calculate the point cloud from the rangemap and metric stamp data values:

//create new metric calibrated cop (with default metric values 1,1,1)
Cvb.Triangulation.Metric3DFactors metric = new Cvb.Triangulation.Metric3DFactors(1, 1, 1);
m_cop = new Cvb.Triangulation.COP3D(m_rangemap, metric);
AddStatusTextToRichBox("COP generated successfully.");

//Create MotionMatrix
float[,] MotionMatrixF = new float[3,4];

MotionMatrixF.SetValue((m_stampData[07] / 1000000f),  0, 0);
MotionMatrixF.SetValue(0,                             0, 1);
MotionMatrixF.SetValue(0,                             0, 2);
        MotionMatrixF.SetValue((m_stampData[06] / 1000000f - 34.5f),  0, 3);  // negative offset of 34.5mm needed, reason unclear!

MotionMatrixF.SetValue(0,                             1, 0);
MotionMatrixF.SetValue((m_stampData[09] / 1000000f),  1, 1);
MotionMatrixF.SetValue(0,                             1, 2);
MotionMatrixF.SetValue((m_stampData[08] / 1000000f + 34.5f),  1, 3);  // positive offset of 34.5mm needed, reason unclear!

MotionMatrixF.SetValue(0,                             2, 0);
MotionMatrixF.SetValue(0,                             2, 1);
MotionMatrixF.SetValue((m_stampData[11] / 1000000f),  2, 2);
MotionMatrixF.SetValue((m_stampData[10] / 1000000f - (m_stampData[11] / 1000000f)*32768),  2, 3);

Cvb.Triangulation.MotionMatrix MotionMatrix = new Cvb.Triangulation.MotionMatrix(MotionMatrixF);
1 Like