Browsing by Author "Kotowski, Krzysztof"

Now showing 1 - 2 of 2
  • No Thumbnail Available
    ItemOpen Access
    Speckle tracking approaches in speckle correlation sensing
    (Cranfield University, 2017-05-02 13:08) Charrett, Tom; Kotowski, Krzysztof; Tatam, Ralph
    Data and code used to generate the conference paper: "Speckle tracking approaches in speckle correlation sensing" Thomas O. H. Charrett, Krzysztof Kotowski, and Ralph P. Tatam SPIE Optics and Optoelectonics, Prague, 2017. Files: ------ lib_feature_tracking.py - python module/library used to simplify the other scripts feature detectors.py - python script used to test processing times of feature detectors. feature descriptors.py - python script used to test processing times of feature descriptors and matching methods modelled shifts.py - python script used to generate figure 1 - accuracy assesment. experimental shifts.py - python script used to compare feature tracking method with cross correlation using real data (figure 2) experimental rotations.py - python script used to test rotation performance using experimental data. Used to generate figure 3. random positions.npy - 100 x (512,512) independent speckle patterns in numpy binary format. Used for table 1, table 2 and figure 1 linear move direction=0.0 speed=5.0mms-1.npy - 100 x (512,512) speckle patterns recorded using a speckle velocimetry sensor on XY stages travelling at 5mm/s in the y-direction. In numpy binary format.Used for figure 2. z rotation.npy - 721 x (512,512) speckle patterns for angles 0 to 360.0 degrees in 0.5 degree steps. Used for figure 3. Comments: ---------------- OpenCV version: 3.1.0 Numpy python library available at http://www.numpy.org/. Numpy version: 1.10.2 Load numpy binary format using: >>> import numpy as np >>> imgs = np.load( filename )
  • Thumbnail Image
    ItemOpen Access
    Speckle tracking approaches in speckle sensing
    (SPIE, 2017-05-16) Charrett, Thomas O. H.; Kotowski, Krzysztof; Tatam, Ralph P.
    This paper reports some initial investigations into the application of feature tracking algorithms as an alternative data processing method for speckle correlation sensors capable of determining both the speckle pattern translation and rotation. The accuracy of translation measurements using the feature tracking approach was found to be similar to that of correlation based processing with accuracies of < 0.04 pixels. Rotation measurement accuracies of< 0.05 ◦ are found to be achievable over angle range ±20 ◦ , limited by the failure to match speckles at larger rotation angles.