Reducing force drift for static load


#1

After a static object was placed on the Morph, 53 frames of data were scanned (~15 seconds). The static object covered ~10% of the Morph. 1,029,525 force data points (53 * 19,425) were captured.

Per the attached figure, the force was plotted to identify any drift, i.e., for cases where a sensor experienced max and min forces that were not equal. In the plot below, the x-axis is the max force; the y-axis is max-min.

Because it was a static load, the data for a perfect device should all lie on the x-axis (max-min = 0). In this example, 1675 sensels drifted, which was all of those under the static object. The rest of the data is at the origin (zero force, zero drift).

The scan sensitivity was high because other settings rendered the footprint far too “fuzzy”.

I would like to explore if the drift can be reduced, especially for the lower force levels for which the relative uncertainty is high.

Thanks!
Brad


#2

Hi Brad, thank you for contacting us. When you captured this data, did you disable dynamic baselining? Dynamic baselining removes some force from the overall image in order to prevent spurious contacts, which could lead to force drift in your measurments. http://guide.sensel.com/sensel_h/index.html#//apple_ref/c/func/senselSetDynamicBaselineEnabled


#3

Yes, dynamic baselining was disabled. Also (FWIW), blob merging is left enabled, else the footprint is rather fragmented.

Relevant code snippet follows:

    senselSetDynamicBaselineEnabled(handle,false);
    senselSetScanDetail(handle,SCAN_DETAIL_HIGH);

    //Set the frame content to scan force data
    senselSetFrameContent(handle, FRAME_CONTENT_PRESSURE_MASK | FRAME_CONTENT_CONTACTS_MASK);
    senselSetContactsMask(handle, CONTACT_MASK_ELLIPSE | CONTACT_MASK_BOUNDING_BOX);