Image Processing

This module contains utilities for image processing.

Functions

block((object)src, (object)dst, …) Performs a block decomposition of a 2D array/image.
crop((object)src, (object)dst, (int)crop_y, …) Crop a 2D or 3D array/image.
draw_box((object)image, (int)x, (int)y, …) Draws a box at the image using the draw_line() primitive.
draw_cross((object)image, (int)x, (int)y, …) Draws a cross with a given radius and color at the image.
draw_cross_plus((object)image, (int)x, …) Draws a cross with a given radius and color at the image.
draw_line((object)image, (int)x1, (int)y1, …) Draws a line between two points p1(x1,y1) and p2(x2,y2).
draw_point((object)image, (int)x, (int)y, …) Draws a point in the given image.
draw_point_((object)image, (int)x, (int)y, …) Draws a point in an image.
extrapolate_mask((object)src_mask, (object)img) Extrapolate a 2D array/image, taking a boolean mask into account.
flip((object)src, (object)dst) Flip a 2D or 3D array/image upside-down.
flop((object)src, (object)dst) Flop a 2D or 3D array/image left-right.
flow_error((object)i1, (object)i2, …) Computes the generalized flow error: E = i2(x-u,y-v) - i1(x,y))
gamma_correction((object)src, (object)dst, …) Performs a power-law gamma correction on a 2D blitz array/image.
get_angle_to_horizontal((float)left_y, …) Get the angle needed to level out (horizontally) two points.
get_block_3d_output_shape((object)src, …) Returns the shape of the output 2D blitz array/image, when calling bob.ip.block() which performs a block decomposition of a 2D array/image, and saving the results in a 3D array (block_index, y, x).
get_block_4d_output_shape((object)src, …) Returns the shape of the output 2D blitz array/image, when calling bob.ip.block() which performs a block decomposition of a 2D array/image, and saving the results in a 4D array (block_index_y, block_index_x, y, x).
get_rotated_output_shape((object)input, …) Returns the shape of the output image when rotating the given input image with the desired angle.
get_shear_x_shape((object)src, (float)shear) Returns the shape of the output 2D array/image, when calling shear_x.
get_shear_y_shape((object)src, (float)shear) Returns the shape of the output 2D array/image, when calling shear_y.
gray_to_rgb((object)y, (object)dtype) Converts a grayscale pixel to RGB by copying the gray value to all 3 bands. Returns a tuple with (r,g,b) values. This method is just here for convenience.
histogram((object)src) Compute an histogram of a 2D array
histogram_((object)src, (object)histo [, …) Compute an histogram of a 2D array.
histogram_equalization((object)src, (object)dst) Computes the histogram equalization of the given src image and fills the dst image.
hog_compute_histogram((object)mag, …) Computes an Histogram of Gradients for a given ‘cell’.
hog_compute_histogram_((object)mag, …) Computes an Histogram of Gradients for a given ‘cell’.
hsl_to_rgb((object)hue, (object)saturation, …) Converts a HSL color-pixel to RGB as defined in http://en.wikipedia.org/wiki/HSL_and_HSV. Returns a tuple with (r,g,b) values.
hsv_to_rgb((object)hue, (object)saturation, …) Converts a HSV color-pixel to RGB as defined in http://en.wikipedia.org/wiki/HSL_and_HSV. Returns a tuple with (r,g,b) values.
integral((object)src, (object)dst [, …) Compute the integral image of a 2D blitz array (image).
laplacian_avg_hs((object)input) An approximation to the Laplacian operator.
laplacian_avg_hs_opencv((object)input) An approximation to the Laplacian (averaging) operator.
max_rect_in_mask((object)arg1) src
normalize_block((object)hist, …) normalizes a set of cells (Histogram of Gradients), and returns the corresponding block descriptor.
normalize_block_((object)hist, …) normalizes a set of cells (Histogram of Gradients), and returns the corresponding block descriptor.
normalize_gabor_jet((object)gabor_jet) Normalizes the Gabor jet (with or without phase) to unit Euclidean length.
rgb_to_gray((object)red, (object)green, …) Converts a RGB color-coded pixel to Grayscale using the CCIR 601 (Kb = 0.114, Kr = 0.299) norm as discussed here: http://www.fourcc.org/fccyvrgb.php.
rgb_to_hsl((object)red, (object)green, …) Converts a RGB color-pixel to HSL as defined in http://en.wikipedia.org/wiki/HSL_and_HSV. Returns a tuple with (h,s,l) values.
rgb_to_hsv((object)red, (object)green, …) Converts a RGB color-pixel to HSV as defined in http://en.wikipedia.org/wiki/HSL_and_HSV. Returns a tuple with (h,s,v) values.
rgb_to_yuv((object)red, (object)green, …) Converts a RGB color-coded pixel to YUV (Y’CbCr) using the CCIR 601 (Kb = 0.114, Kr = 0.299) norm as discussed here: http://en.wikipedia.org/wiki/YCbCr and here: http://www.fourcc.org/fccyvrgb.php.
rotate((object)input, (object)output, …) Rotates the given input image into the given output image.
scale((object)src, (object)dst [, …) Scales an image to the dimensions given by the allocated destination image.
scale_as((object)src, (float)scaling_factor) Scales an image according to the provided scaling factor.
shear_x((object)src, (object)dst, …) Shears a 2D array/image with the given shear parameter along the X-dimension.
shear_y((object)src, (object)dst, …) Shears a 2D array/image with the given shear parameter along the Y-dimension.
shift((object)src, (object)dst, …) Shift a 2D or 3D array/image.
try_draw_point((object)image, (int)x, …) Tries to draw a point at the given image.
yuv_to_rgb((object)y, (object)u, (object)v, …) Converts a YUV (Y’CbCr) color-coded pixel to RGB using the CCIR 601 (Kb = 0.114, Kr = 0.299) norm as discussed here: http://en.wikipedia.org/wiki/YCbCr and here: http://www.fourcc.org/fccyvrgb.php.
zigzag((object)src, (object)obj [, …) Extracts a 1D NumPy array using a zigzag pattern from a 2D array/image.

Classes

BlockNorm
CentralGradient((object)self, …) This class computes the spatio-temporal gradient using a 3-term approximation composed of 2 separable kernels (one for the diference term and another one for the averaging term).
DCTFeatures((object)self, (int)block_h, …) Objects of this class, after configuration, can extract DCT features as described in the following reference.
ELBPType Different types of LBP codes
FaceEyesNorm((object)self, …) Objects of this class, after configuration, can extract and normalize faces, given their eye center coordinates.
ForwardGradient((object)self, …) This class computes the spatio-temporal gradient using a 2-term approximation composed of 2 separable kernels (one for the diference term and another one for the averaging term).
GLCM(dtype[, num_levels, min_level, …]) Objects of this class, after configuration, can calculate GLCM matrix on a given input image.
GLCMProp((object)self) Objects of this class, after configuration, can compute a number of texture properties on a Grey-Level Co-occurence Matrix (GLCM).
GSSKeypoint((object)self, (float)scale, …) Structure to describe a keypoint on the Gaussian Scale Space.
GSSKeypointInfo((object)self [, …) This is a companion structure to the bob.ip.GSSKeypoin.
GaborKernel((object)self, …) This class can be used to filter an image with a single Gabor wavelet.
GaborWaveletTransform((object)self [, …) This class can be used to perform a Gabor wavelet transform from one image to an image of (normalized) Gabor jets or to a complex-valued multi-layer trafo image.
Gaussian((object)self [, (int)radius_y=1 [, …) This class allows after configuration to perform gaussian smoothing.
GaussianScaleSpace((object)arg1, …) This class allows after configuration the generation of Gaussian Pyramids that can be used to extract SIFT features.
GeomNorm((object)self, …) Objects of this class, after configuration, can perform a geometric normalization.
GradientMagnitudeType
GradientMaps((object)self, (int)height, …) Objects of this class, after configuration, can extract gradient magnitude and orientation maps.
HOG((object)self, (int)height, (int)width [, …) Objects of this class, after configuration, can extract Histogram of Gradients (HOG) descriptors.
HornAndSchunckFlow((object)self, (object)shape) This is a clone of the Vanilla HornAndSchunck method that uses a Sobel gradient estimator instead of the forward estimator used by the classical method.
HornAndSchunckGradient((object)self, …) This class computes the spatio-temporal gradient using the same approximation as the one described by Horn & Schunck in the paper titled ‘Determining Optical Flow’, published in 1981, Artificial Intelligence, * Vol.
IsotropicGradient((object)self, (object)shape) This class computes the spatio-temporal gradient using a isotropic filter.
LBP((object)self, (int)neighbors, …) A class for the LBP operators
LBPHSFeatures((object)self, (int)block_h, …) Constructs a new LBPHSFeatures object to extract histogram of LBP over 2D blitz arrays/images.
LBPTop((object)self, (LBP)xy, (LBP)xt, (LBP)yt) Constructs a new LBPTop object starting from the algorithm configuration.
Median_float64((object)self, (int)radius_y, …) Objects of this class, after configuration, can perform a median filtering operation.
Median_uint16((object)self, (int)radius_y, …) Objects of this class, after configuration, can perform a median filtering operation.
Median_uint8((object)self, (int)radius_y, …) Objects of this class, after configuration, can perform a median filtering operation.
MultiscaleRetinex((object)self [, …) This class allows after configuration to apply the Self Quotient Image algorithm to images.
PrewittGradient((object)self, (object)shape) This class computes the spatio-temporal gradient using a 3-D sobel filter.
RescaleAlgorithm
RotateAlgorithm
SIFT((object)self, (int)height, (int)width, …) This class allows after configuration the extraction of SIFT descriptors.
SelfQuotientImage((object)self [, …) This class allows after configuration to apply the Self Quotient Image algorithm to images.
Sobel((object)self [, …) This class allows after configuration to perform Sobel filtering.
SobelGradient((object)self, (object)shape) This class computes the spatio-temporal gradient using a 3-D sobel filter.
TanTriggs((object)self [, …) Objects of this class, after configuration, can preprocess images.
VLDSIFT((object)self, (int)height, …) Computes dense SIFT features using the VLFeat library
VLSIFT((object)self, (int)height, …) Computes SIFT features using the VLFeat library
VanillaHornAndSchunckFlow((object)self, …) Calculates the Optical Flow between two sequences of images (i1, the starting image and i2, the final image).
WeightedGaussian((object)self [, …) This class performs weighted gaussian smoothing (anisotropic filtering).