v0.8.7

• Switch to official release of Aparapi at version 3.0.2
• Update build instructions
• Update copyright information

v0.8.6

Fixes Bug in:

• GaussianFilter2D where the filter was improperly using the image height as its width, resulting in buggy behavior for asymmetric resolutions.

Other changes:

• Switched default Lagrange multiplier value of Liu-Shen from 1000 to 4.
• Updated the About dialog box information.

v0.8.5

Fixes bug in:

• dense Liu-Shen Aparapi HPC/GPU implementations to work with small Lagrange multiplier values of around 4.0
• dense Liu-Shen Aparapi HPC/GPU implementations to properly handle synchronization through Java and OpenCL atomics
• Liu-Shen Java CPU implementation to work with small Lagrange multiplier values of around 4.0
• Lucas-Kanade implementations behavior at the image top pixel row
• Image.clipImage() method when clipping images to targets that need instancing

Simplifies:

• Lucas-Kanade Aparapi HPC/GPU implementations around readPixelWithWarp(...) method

v0.8.4

QuickLabPIV-ng v0.8.4

• Fixes GPU FFT Benchmark diagnostic test
• Fixes missing logo in AboutDialog window
• Updated AboutDialog window co-supervisors list
• All unit tests passing

v0.8.3

v0.8.3 - Initial release of QuickLabPIV-ng

• PIV and Hybrid PIV high performance computing application with GP-GPU OpenCL support (by Aparapi)
• Friendly Graphical User Interface (GUI)
• Supports both dense and sparse Liu-Shen combined with Lucas-Kanade and Lucas-Kanade only Optical Flow methods
• Support classic PIV and PIV with warping modes
• MATLAB file format data export including multi-volume support
• Adaptive PIV support with configurable start and end Interrogation Area window sizes
• Multiple sub-pixel methods including Polynomial Gaussian 1D-1D and Hongwei Guo's Gaussian 1D-1D Robust Linear regression, among others
• Vector validation and substitution including secondary peak substitution
• Single XML configuration file
• Support images sequences and image pairs
• Selectable GPU per CPU threads distribution