View Issue Details [ Jump to Notes ] | [ Print ] | ||||||||||||
ID | Project | Category | View Status | Date Submitted | Last Update | ||||||||
0011481 | ITK | ITK | public | 2010-11-18 23:03 | 2010-11-19 09:30 | ||||||||
Reporter | David Doria | ||||||||||||
Assigned To | Gabe Hart | ||||||||||||
Priority | normal | Severity | minor | Reproducibility | sometimes | ||||||||
Status | assigned | Resolution | open | ||||||||||
Platform | Linux | OS | Fedora | OS Version | 13 | ||||||||
Product Version | ITK-4-A3 | ||||||||||||
Target Version | Fixed in Version | ||||||||||||
Summary | 0011481: itkImageToHistogramFilter Strange Behavior when image has zero variance in at least one component | ||||||||||||
Description | If at least one component of an image has zero variance, iterating over the histogram displays strange behavior. How should this case be handled? Throw an error/warning that the bounds do not make sense to compute? Choose arbitrary bin end points (0 - 1, etc)? | ||||||||||||
Steps To Reproduce | I have attached Test.cxx which demonstrates the problem. The CreateImage function creates a mostly black image with one red and one green pixel. There is code to create a blue pixel, but it is commented. If this blue pixel is added, then no components have zero variance and everything works as expected. Without this blue pixel, calling: << histogram->GetIndex(histogramIterator.GetMeasurementVector()) << " " << histogram->GetIndex(histogramIterator.GetInstanceIdentifier()) << " " produces the wrong values for the index. However, calling << histogram->GetIndex(histogramIterator.GetInstanceIdentifier()) << " " << histogram->GetIndex(histogramIterator.GetMeasurementVector()) << " " produces the correct values (note that these are the same calls, just in the opposite order). By "wrong values", I mean the index is actually out of bounds. Here I have created a 2x2x2 histogram, and sometimes the a component of the index is 2, which is invalid. Also, the iterator doesn't seem to traverse the histogram in the correct order. This is probably due to the fact that the position is being computed using invalid bounds (in the zero variance case only). | ||||||||||||
Tags | No tags attached. | ||||||||||||
Resolution Date | |||||||||||||
Sprint | |||||||||||||
Sprint Status | backlog | ||||||||||||
Attached Files | ![]() | ||||||||||||
Relationships | |
Relationships |
Notes | |
There are no notes attached to this issue. |
Notes |
Issue History | |||
Date Modified | Username | Field | Change |
2010-11-18 23:03 | David Doria | New Issue | |
2010-11-18 23:03 | David Doria | Status | new => assigned |
2010-11-18 23:03 | David Doria | Assigned To | => Luis Ibanez |
2010-11-18 23:03 | David Doria | File Added: Test.cxx | |
2010-11-19 09:30 | Gabe Hart | Assigned To | Luis Ibanez => Gabe Hart |
Issue History |
Copyright © 2000 - 2018 MantisBT Team |