A collection of python functions for feature extraction. The features are calculated inside a Region of Interest (ROI) and not for the whole image: the image is actually a polygon! More and more features will be added. Please feel free to point out any mistakes or improvemets. The aim is to create a library for image feature extraction. Message me for more details.
- First Order Statistics/Statistical Features (FOS/SF)
- Gray Level Co-occurence Matrix (GLCM/SGLDM)
- Gray Level Difference Statistics (GLDS)
- Neighborhood Gray Tone Difference Matrix (NGTDM)
- Statistical Feature Matrix (SFM)
- Law's Texture Energy Measures (LTE/TEM)
- Fractal Dimension Texture Analysis (FDTA)
- Gray Level Run Length Matrix (GLRLM)
- Fourier Power Spectrum (FPS)
- Shape Parameters
- Gray Level Size Zone Matrix (GLSZM)
- Higher Order Spectra (HOS)
- Local Binary Pattern (LPB)
- Grayscale Morphological Analysis
- Multilevel Binary Morphological Analysis
- Histogram
- Multi-region histogram
- Correlogram
- Fractal Dimension Texture Analysis (FDTA)
- Amplitude Modulation – Frequency Modulation (AM-FM)
- Discrete Wavelet Transform (DWT)
- Stationary Wavelet Transform (SWT)
- Wavelet Packets (WP)
- Gabor Transform (GT)
- Zernikes’ Moments
- Hu’s Moments
- Threshold Adjacency Matrix (TAS)
- Histogram of Oriented Gradients (HOG)
Download the folder features, add to path and call
from features import *For the following sections, assume
- f is a grayscale image as a numpy ndarray,
- mask is an image as a numpy ndarray but with 2 values: 0 (zero) and 1 (one) with 1 indicating the ROI, where the features shall be calculated (values outside ROI are ignored),
- perimeter is like mask but indicates the perimeter of the ROI. The demo has an analytic way on how to create mask and perimeter, given a set of coordinates.
features, labels = fos(f, mask)
'''
Parameters
----------
f : numpy ndarray
Image of dimensions N1 x N2.
mask : numpy ndarray
Mask image N1 x N2 with 1 if pixels belongs to ROI, 0 else. Give None
if you want to consider ROI the whole image.
Returns
-------
features : numpy ndarray
Feature set as defined in theory.
labels : list
Labels of features.
'''features_mean, features_range, labels_mean, labels_range = glcm_features(f, ignore_zeros)
'''
Parameters
----------
f : numpy ndarray
Image of dimensions N1 x N2.
ignore_zeros : int, optional
Ignore zeros in image f. The default is True.
Returns
-------
features_mean : numpy ndarray
Haralick's features, mean
features_range : numpy ndarray
Haralick's features, same as before but range
labels_mean : list
Labels of features_mean.
labels_range: list
Labels of features_range.
'''features, labels = glds_features(f, mask, Dx, Dy)
'''
Parameters
----------
f : numpy ndarray
Image of dimensions N1 x N2.
mask : numpy ndarray
Mask image N1 x N2 with 1 if pixels belongs to ROI, 0 else.
Dx : int, optional
Array with X-coordinates of vectors denoting orientation. The default
is [0,1,1,1].
Dy : int, optional
Array with Y-coordinates of vectors denoting orientation. The default
is [1,1,0,-1].
Returns
-------
features : numpy ndarray
Feature set as defined in theory.
labels : list
Labels of features.
'''features, labels = ngtdm_features(f, mask, d)
'''
Parameters
----------
f : numpy ndarray
Image of dimensions N1 x N2.
mask : numpy ndarray
Mask image N1 x N2 with 1 if pixels belongs to ROI, 0 else. Give None
if you want to consider ROI the whole image.
d : int, optional
Distance for NGTDM. Default is 1.
Returns
-------
features : numpy ndarray
Feature set as defined in theory.
labels : list
Labels of features.
'''features, labels = sfm_features(f, mask, Lr, Lc)
'''
Parameters
----------
f : numpy ndarray
Image of dimensions N1 x N2.
mask : numpy ndarray
Mask image N1 x N2 with 1 if pixels belongs to ROI, 0 else. Give None
if you want to consider ROI the whole image.
Lr : int, optional
Parameters of SFM. The default is 4.
Lc : int, optional
Parameters of SFM. The default is 4.
Returns
-------
features : numpy ndarray
Feature set as defined in theory.
labels : list
Labels of features.
'''features, labels = lte_measures(f, mask, l)
'''
Parameters
----------
f : numpy ndarray
Image of dimensions N1 x N2.
mask : numpy ndarray
Mask image N1 x N2 with 1 if pixels belongs to ROI, 0 else. Give None
if you want to consider ROI the whole image.
l : int, optional
Law's mask size. The default is 7.
Returns
-------
features : numpy ndarray
Feature set as defined in theory.
labels : list
Labels of features.
'''h, labels = fdta(f, mask, s)
'''
Parameters
----------
f : numpy ndarray
Image of dimensions N1 x N2.
mask : numpy ndarray
Mask image N1 x N2 with 1 if pixels belongs to ROI, 0 else. Give None
if you want to consider ROI the whole image.
s : int, optional
max resolution to calculate Hurst coefficients. The default is 3
Returns
-------
h : numpy ndarray
Hurst coefficients.
labels : list
Labels of h.
'''features, labels = glrlm_features(f, mask, Ng)
'''
Parameters
----------
f : numpy ndarray
Image of dimensions N1 x N2.
mask : numpy ndarray
Mask image N1 x N2 with 1 if pixels belongs to ROI, 0 else. Give None
if you want to consider ROI the whole image.
Ng : int, optional
Image number of gray values. The default is 256.
Returns
-------
features : numpy ndarray
Feature set as defined in theory.
labels : list
Labels of features.
'''features, labels = fps(f, mask)
'''
Parameters
----------
f : numpy ndarray
Image of dimensions N1 x N2.
mask : numpy ndarray
Mask image N1 x N2 with 1 if pixels belongs to ROI, 0 else. Give None
if you want to consider ROI the whole image.
Returns
-------
features : numpy ndarray
Feature set as defined in theory.
labels : list
Labels of features.
'''features, labels = shape_parameters(f, mask, perimeter, pixels_per_mm2)
'''
Parameters
----------
f : numpy ndarray
Image of dimensions N1 x N2.
mask : numpy ndarray
Mask image N1 x N2 with 1 if pixels belongs to ROI, 0 else. Give None
if you want to consider ROI the whole image.
perimeter : numpy ndarray
Image N1 x N2 with 1 if pixels belongs to perimeter of ROI, 0 else.
pixels_per_mm2 : int, optional
Density of image f. The default is 1.
Returns
-------
features : numpy ndarray
Feature set as defined in theory.
labels : list
Labels of features.
'''features, labels = glszm_features(f, mask)
'''
Parameters
----------
f : numpy ndarray
Image of dimensions N1 x N2.
mask : numpy ndarray
Mask image N1 x N2 with 1 if pixels belongs to ROI, 0 else.
Returns
-------
features : numpy ndarray
Feature set as defined in theory.
labels : list
Labels of features.
'''features, labels = hos_features(f, th)
'''
Parameters
----------
f : numpy ndarray
Image of dimensions N1 x N2.
th : list, optional
Angle to calculate Radon Transform. The default is [135,140].
Returns
-------
features : numpy ndarray
Entropy of bispectrum of radeon transform of image for each angle in theta.
labels : list
Labels of features.
'''features, labels = lbp_features(f, mask, P, R)
'''
Parameters
----------
f : numpy ndarray
Image of dimensions N1 x N2.
mask : numpy ndarray
Mask image N1 x N2 with 1 if pixels belongs to ROI, 0 else. Give None
if you want to consider ROI the whole image.
P : list, optional
Number of points in neighborhood. The default is [8,16,24].
R : list, optional
Radius/Radii. The default is [1,2,3].
Returns
-------
features : numpy ndarray
Energy and entropy of LBP image (2 x 1).
labels : list
Labels of features.
'''pdf, cdf = grayscale_morphology_features(f, N)
'''
Parameters
----------
f : numpy ndarray
Image of dimensions N1 x N2.
N : np.array, optional
Maximum number of scales. The default is 30.
Returns
-------
pdf : numpy ndarray
Probability density function (pdf) of pattern spectrum.
cdf : numpy ndarray
Cumulative density function (cdf) of pattern spectrum.
'''pdf_L, pdf_M, pdf_H, cdf_L, cdf_M, cdf_H = multilevel_binary_morphology_features(img, mask, N, thresholds):
'''
Parameters
----------
img : numpy ndarray
Image of dimensions N1 x N2.
mask : numpy ndarray
Mask image N1 x N2 with 1 if pixels belongs to ROI, 0 else. Give None
if you want to consider ROI the whole image.
N : np.array, optional
Maximum number of scales. The default is 30.
thresholds: list, optional
Thresholds to get the 3 binary images. The default is [25, 50].
Returns
-------
pdf_L : numpy ndarray
Probability density function (pdf) of pattern spectrum for image L.
pdf_M : numpy ndarray
Probability density function (pdf) of pattern spectrum for image M.
pdf_H : numpy ndarray
Probability density function (pdf) of pattern spectrum for image H.
cdf_L : numpy ndarray
Cumulative density function (cdf) of pattern spectrum for image L.
cdf_M : numpy ndarray
Cumulative density function (cdf) of pattern spectrum for image M.
cdf_H : numpy ndarray
Cumulative density function (cdf) of pattern spectrum for image H.
''' H, labels = histogram(f, mask, bins)
'''
Parameters
----------
f : numpy ndarray
Image of dimensions N1 x N2.
mask : numpy ndarray
Mask image N1 x N2 with 1 if pixels belongs to ROI, 0 else. Give None
if you want to consider ROI the whole image.
bins : int, optional
Bins for histogram. The default is 32.
Returns
-------
H : numpy ndarray
Histogram of image f for 256 gray levels.
labels : list
Labels of features, which are the bins' number.
'''features, labels = multiregion_histogram(f, mask, bins, num_eros,
57AE
square_size)
'''
Parameters
----------
f : numpy ndarray
Image of dimensions N1 x N2.
mask : numpy ndarray
Mask image N1 x N2 with 1 if pixels belongs to ROI, 0 else. Give None
if you want to consider ROI the whole image.
bins : int, optional
Bins for histogram. Default is 32.
num_eros : int, optional
Times of erosion to be performed. Default is 3.
square_size : int, optional
Kernel where erosion is performed, here a squared. Default square's
size is 3.
Returns
-------
features : numpy ndarray
Histogram of f and f after erosions as a vector e.g. [32 x num_eros].
labels : list
Labels of features.
'''Hd, Ht, labels = correlogram(f, mask, bins_digitize, bins_hist, flatten)
'''
Parameters
----------
f : numpy ndarray
Image of dimensions N1 x N2.
mask : numpy ndarray
Mask image N1 x N2 with 1 if pixels belongs to ROI, 0 else. Give None
if you want to consider ROI the whole image.
bins_digitize : int, optional
Number of bins for discrete distances and thetas. The default is 32.
bins_hist : int, optional
Number of bins for histogram. The default is 32.
flatten : bool, optional
Return correlogram as 1d array if True or 2d array if False. The
default is False.
Returns
-------
Hd : numpy ndarray
Correlogram for distance.
Ht : numpy ndarray
Correlogram for angles.
labels : list
Labels of features.
'''h, labels = fdta(f, mask, s)
'''
Parameters
----------
f : numpy ndarray
Image of dimensions N1 x N2.
mask : numpy ndarray
Mask image N1 x N2 with 1 if pixels belongs to ROI, 0 else. Give None
if you want to consider ROI the whole image.
s : int, optional
max resolution to calculate Hurst coefficients. The default is 3
Returns
-------
h : numpy ndarray
Hurst coefficients.
labels : list
Labels of h.
'''features, labels = amfm_features(f, bins)
'''
Parameters
----------
f : numpy ndarray
Image of dimensions N1 x N2.
bins: int, optional
Bins for the calculated histogram. The default is 32.
Returns
-------
features : numpy ndarray
Histogram of IA, IP, IFx, IFy as a concatenated vector.
labels : list
Labels of features.
'''features, labels = dwt_features(f, mask, wavelet, levels)
'''
Parameters
----------
f : numpy ndarray
Image of dimensions N1 x N2.
mask : numpy ndarray
Mask image N1 x N2 with 1 if pixels belongs to ROI, 0 else. Give None
if you want to consider ROI the whole image.
wavelet : str, optional
Filter to be used. Check pywt for filter families. The default is 'bior3.3'
levels : int, optional
Levels of decomposition. Default is 3.
Returns
-------
features : numpy ndarray
Mean and std of each detail image. Appromimation images are ignored.
labels : list
Labels of features.
'''features, labels = swt_features(f, mask, wavelet, levels)
'''
Parameters
----------
f : numpy ndarray
Image of dimensions N1 x N2.
mask : numpy ndarray
Mask image N1 x N2 with 1 if pixels belongs to ROI, 0 else. Give None
if you want to consider ROI the whole image.
wavelet : str, optional
Filter to be used. Check pywt for filter families. The default is 'bior3.3'
levels : int, optional
Levels of decomposition. Default is 3.
Returns
-------
features : numpy ndarray
Mean and std of each detail image. Appromimation images are ignored.
labels : list
Labels of features.
'''features, labels = wp_features(f, mask, wavelet, maxlevel)
'''
Parameters
----------
f : numpy ndarray
Image of dimensions N1 x N2.
mask : numpy ndarray
Mask image N1 x N2 with 1 if pixels belongs to ROI, 0 else. Give None
if you want to consider ROI the whole image.
wavelet : str, optional
Filter to be used. Check pywt for filter families. The default is 'cof1'
maxlevel : int, optional
Levels of decomposition. Default is 3.
Returns
-------
features : numpy ndarray
Mean and std of each detail image. Appromimation images are ignored.
labels : list
Labels of features.
'''features, labels = gt_features(f, mask, deg, freq)
'''
Parameters
----------
f : numpy ndarray
Image of dimensions N1 x N2.
mask : numpy ndarray
Mask image N1 x N2 with 1 if pixels belongs to ROI, 0 else. Give None
if you want to consider ROI the whole image.
deg: int, optinal
Quantized degrees. The default is 4 (0, 45, 90, 135 degrees)
freq: list, optional
frequency of the gabor kernel. The default is [0.05, 0.4]
Returns
-------
features : numpy ndarray
Mean and std for the resulted image: (f o gabor_filter)(x,y)
labels : list
Labels of features.
''' features, labels = zernikes_moments(f, radius)
'''
Parameters
----------
f : numpy ndarray
Image of dimensions N1 x N2.
radius : int, optional
Radius to calculate Zernikes moments. The default is 9.
Returns
-------
features : numpy ndarray
Zernikes' moments.
labels : list
Labels of features.
'''features, labels = hu_moments(f)
'''
Parameters
----------
f : numpy ndarray
Image of dimensions N1 x N2.
Returns
-------
features : numpy ndarray
Hu's moments.
labels : list
Labels of features.
'''features, labels = tas_features(f)
'''
Parameters
----------
f : numpy ndarray
Image of dimensions N1 x N2.
Returns
-------
features : numpy ndarray
Feature values.
labels : list
Labels of features.
''' fd, labels = hog_features(f, ppc, cpb)
'''
Parameters
----------
f : numpy ndarray
Image of dimensions N1 x N2.
ppc : int, optional
Pixels per cell. The default is 8.
cpb : int, optional
Cells per block. The default is 3.
Returns
-------
fd : numpy ndarray
Histogram of Oriented Gradients flattened.
labels : list
Labels of features.
'''Reach out to me:
- Bradski, G. (2000). The OpenCV Library. Dr. Dobb's Journal of Software Tools.
- Coelho, L.P. 2013. Mahotas: Open source software for scriptable computer vision. Journal of Open Research Software 1(1):e3, DOI: http://dx.doi.org/10.5334/jors.ac
- Hunter, J. D. (2007). Matplotlib: A 2D graphics environment. Computing in Science & Engineering, 9(3), 90–95.
- Harris, C.R., Millman, K.J., van der Walt, S.J. et al. Array programming with NumPy. Nature 585, 357–362 (2020). DOI: 0.1038/s41586-020-2649-2. (Publisher link).
- Gregory R. Lee, Ralf Gommers, Filip Wasilewski, Kai Wohlfahrt, Aaron O’Leary (2019). PyWavelets: A Python package for wavelet analysis. Journal of Open Source Software, 4(36), 1237, https://doi.org/10.21105/joss.01237.
- Virtanen, P., Gommers, R., Oliphant, T. E., Haberland, M., Reddy, T., Cournapeau, D., … SciPy 1.0 Contributors. (2020). SciPy 1.0: Fundamental Algorithms for Scientific Computing in Python. Nature Methods, 17, 261–272. https://doi.org/10.1038/s41592-019-0686-2
- Van der Walt, S., Sch"onberger, Johannes L, Nunez-Iglesias, J., Boulogne, Franccois, Warner, J. D., Yager, N., … Yu, T. (2014). scikit-image: image processing in Python. PeerJ, 2, e453.
- Acharya U, R., Chua, C. K., Ng, E. Y., Yu, W., & Chee, C. (2008, 12 01). Application of Higher Order Spectra for the Identification of Diabetes Retinopathy Stages. Journal of Medical Systems, 32, 481-488. doi:10.1007/s10916-008-9154-8
- Acharya, U. R., Chua, K., Lim, T.-C., Tay, D., & Suri, J. (2009, 12). Automatic identification of epileptic EEG signals using nonlinear parameters. Journal of Mechanics in Medicine and Biology, 9, 539-553. doi:10.1142/S0219519409003152
- Amadasun, M., & King, R. (1989). Textural features corresponding to textural properties. IEEE Trans. Syst. Man Cybern., 19, 1264-1274.
- Chua, K., Chandran, V., Acharya, U. R., & Lim, C. (2009, 6). Automatic identification of epileptic electroencephalography signals using higher-order spectra. Proceedings of the Institution of Mechanical Engineers. Part H, Journal of engineering in medicine, 223, 485-95. doi:10.1243/09544119JEIM484
- Chua, K., Chandran, V., Acharya, U. R., & Lim, C. (2011, 12). Application of Higher Order Spectra to Identify Epileptic EEG. Journal of medical systems, 35, 1563-71. doi:10.1007/s10916-010-9433-z
- Galloway, M. M. (1975). Texture analysis using gray level run lengths. Computer Graphics and Image Processing, 4, 172-179. doi:https://doi.org/10.1016/S0146-664X(75)80008-6
- Haralick, R., Shanmugam, K., & Dinstein, I. (1973, 1). Textural Features for Image Classification. IEEE Trans Syst Man Cybern, SMC-3, 610-621.
- Hu, M. (1962). Visual pattern recognition by moment invariants. IRE Trans. Inf. Theory, 8, 179-187.
- Laws, K. (1980). Rapid texture identification.
- Lee, G. R., Gommers, R., Waselewski, F., Wohlfahrt, K., O&, A., #8217, & Leary. (2019). PyWavelets: A Python package for wavelet analysis. Journal of Open Source Software, 4, 1237. doi:10.21105/joss.01237
- Liu, M., He, Y., & Ye, B. (2007). Image Zernike moments shape feature evaluation based on image reconstruction. Geo-spatial Information Science, 10, 191-195. doi:10.1007/s11806-007-0060-x
- Mandelbrot, B. (1977). Fractal Geometry of Nature.
- Maragos, P. (1989, 8). Pattern spectrum and multiscale shape representation. IEEE Trans Pattern Anal Mach Intell. Pattern Analysis and Machine Intelligence, IEEE Transactions on, 11, 701-716. doi:10.1109/34.192465
- Maragos, P., & Ziff, R. (1990, 6). Threshold Superposition in Morphological Image Analysis Systems. Pattern Analysis and Machine Intelligence, IEEE Transactions on, 12, 498-504. doi:10.1109/34.55110
- Murray Herrera, V. M. (2009). AM-FM methods for image and video processing. Retrieved from https://digitalrepository.unm.edu/
- Ojala, T., Pietikäinen, M., & Harwood, D. (1996). A comparative study of texture measures with classification based on featured distributions. Pattern Recognit., 29, 51-59.
- Ojala, T., Pietikäinen, M., & Maenpaa, T. (2002, 8). Multiresolution Gray-Scale and Rotation Invariant Texture Classification with Local Binary Patterns. Pattern Analysis and Machine Intelligence, IEEE Transactions on, 24, 971-987. doi:10.1109/TPAMI.2002.1017623
- Teague, M. R. (1980, 8). Image analysis via the general theory of moments∗. J. Opt. Soc. Am., 70, 920–930. doi:10.1364/JOSA.70.000920
- Thibault, G., Fertil, B., Navarro, C., Pereira, S., Cau, P., Lévy, N., . . . Mari, J. (2009). Texture indexes and gray level size zone matrix. Application to cell nuclei classification.
- Toet, A. (1990). A hierarchical morphological image decomposition. Pattern Recognition Letters, 11, 267-274. doi:https://doi.org/10.1016/0167-8655(90)90065-A
- Tsiaparas, N., Golemati, S., Andreadis, I., Stoitsis, J. S., Valavanis, I., & Nikita, K. S. (2011, 1). Comparison of Multiresolution Features for Texture Classification of Carotid Atherosclerosis From B-Mode Ultrasound. IEEE Transactions on Information Technology in Biomedicine, 15, 130–137. doi:10.1109/titb.2010.2091511
- Weszka, J. S., Dyer, C. R., & Rosenfeld, A. (1976). A Comparative Study of Texture Measures for Terrain Classification. IEEE Transactions on Systems, Man, and Cybernetics, SMC-6, 269-285. doi:10.1109/TSMC.1976.5408777
- Wu, C.-M., & Chen, Y.-C. (1992). Statistical feature matrix for texture analysis. CVGIP: Graphical Models and Image Processing, 54, 407-419. doi:https://doi.org/10.1016/1049-9652(92)90025-S
- Wu, C.-M., Chen, Y.-C., & Hsieh, K.-S. (1992). Texture features for classification of ultrasonic liver images. IEEE Transactions on Medical Imaging, 11, 141-152. doi:10.1109/42.141636