Welcome to pyaesthetics’s documentation! - V. 0.0.7¶
Analysis¶
Brightness analysis¶
This module contains function to evaluate the brightness of an image. It includes a converter for sRGB to RGB, evaluation of relative luminance according to BT.709 and BT.601
@author: Giulio Gabrieli
-
brightness.relativeLuminance_BT601(img)¶ This function evaluates the brightness of an image by mean of Y, where Y is evaluated as:
Y = 0.587G + 0.114B + 0.299R B = mean(Y)
Parameters: img (numpy.ndarray) – image to analyze, in RGB Returns: mean brightness Return type: float
-
brightness.relativeLuminance_BT709(img)¶ This function evaluates the brightness of an image by mean of Y, where Y is evaluated as:
Y = 0.7152G + 0.0722B + 0.2126R B = mean(Y)
Parameters: img (numpy.ndarray) – image to analyze, in RGB Returns: mean brightness Return type: float
-
brightness.sRGB2RGB(img)¶ this function converts a sRGB img to linear RGB values.
It loops through each pixel, and apply a conversion to pass from sRGB to linear RGB value.
Parameters: img (numpy.ndarray) – image to analyze, in sRGB Returns: image to analyze, in RGB Rtyipe: numpy.ndarray
Colorfulness analysis¶
This module contains function to evaluate the colorfulness of an image in both the HSV and RGB color spaces.
@author: Giulio Gabrieli
-
colorfulness.colorfulnessHSV(img)¶ This function evaluates the colorfulness of a picture using the formula described in Yendrikhovskij et al., 1998. Input image is first converted to the HSV color space, then the S values are selected. Ci is evaluated with a sum of the mean S and its std, as in:
Ci = mean(Si)+ std(Si)
Parameters: img (numpy.ndarray) – image to analyze, in RGB Returns: colorfulness index Return type: float
-
colorfulness.colorfulnessRGB(img)¶ This function evaluates the colorfulness of a picture using Metric 3 described in Hasler & Suesstrunk, 2003. Ci is evaluated with as:
Ci =std(rgyb) + 0.3 mean(rgyb) [Equation Y] std(rgyb) = sqrt(std(rg)^2+std(yb)^2) mean(rgyb) = sqrt(mean(rg)^2+mean(yb)^2) rg = R - G yb = 0.5(R+G) - B
Parameters: img (numpy.ndarray) – image to analyze, in RGB Returns: colorfulness index Return type: float
-
colorfulness.sRGB2RGB(img)¶ this function converts a sRGB img to linear RGB values.
Parameters: img (numpy.ndarray) – image to analyze, in sRGB Returns: image to analyze, in RGB Rtyipe: numpy.ndarray
Color Detection¶
This module contains function to evaluate the presence of different colors of an image. It uses the 16 basic colors defined in the W3C specifications.
@author: Giulio Gabrieli
-
colorDetection.getColorsW3C(img, plot=False)¶ This functions is used to get a simplified color palette (W3C siteens basic colors).
F = 255 C0 = 192 80 = 128
Parameters: - img (numpy.ndarray) – image to analyze in RGB
- plot (boolean) – whether to plot or not the results
Returns: percentage distribution of colors according to the W3C sixteens basic colors
Return type: list of shape 16x2, where x[0] is the color name and x[1] the percentage of pixels most similar to that color in the image
Face Detection¶
This is an entrypoint for automatic analysis of a website.
Created on Mon Apr 16 22:40:46 2018
@author: giulio
-
faceDetection.getFaces(img, plot=False)¶ This functions uses CV2 to get the faces in a pciture.
Parameters: - img (numpy.ndarray) – image to analyze in RGB
- plot (boolean) – whether to plot or not the results
QuadTree Decomposition analysis¶
This file contains class and functions to perform a Quadratic Tree decomposition of an image and to visually inspect it.
Created on Mon Apr 16 11:49:45 2018
@author: giulio
-
class
quadTreeDecomposition.quadTree(img, minStd, minSize)¶ This class is used to perfrom a QuadTree decomposition of an image.
During initialization, QuadTree decomposition is done and result are store in self.blocks as a list containing [x,y,height, width,Std].
To visualize the results, use plot().
-
plot(edgecolor='red', facecolor='none', linewidth=1)¶ This function is used to generate a graphical representation of the QuadTree decomposition.
Parameters: - edgecolor (string) – color of the rectangles, default is red
- facecolor (string) – color used for rectangles fills. Default is none.
- linewidth – width in px of the rectangles’ borders. Default is 1.
Returns: plot with image and leaves of the quadTree Decomposition
-
quadTreeDecomposition(img, x, y, minStd, minSize)¶ This function evaluate the mean and std of an image, and decides Whether to perform or not other 2 splits of the leave.
Parameters: - img (numpy.ndarray) – img to analyze
- x (int) – x offset of the leaves to analyze
- Y (int) – Y offset of the leaves to analyze
MinStd: Std threshold for subsequent splitting
MinSize: Size threshold for subsequent splitting, in pixel
-
Symmetry analysis¶
This module contains functions to compute the degree of symmetry of an image. - Symmetry by QuadTree Decomposition
Created on Mon Apr 16 11:49:45 2018
@author: giulio
-
symmetry.getSymmetry(img, minStd, minSize, plot=False)¶ This function returns the degree of symmetry (0-100) between the left and right side of an image
Parameters: img (numpy.ndarray) – img to analyze MinStd: Std threshold for subsequent splitting MinSize: Size threshold for subsequent splitting, in pixel Returns: degree of vertical symmetry Return type: float
-
class
symmetry.quadTree(img, minStd, minSize)¶ This class is used to perfrom a QuadTree decomposition of an image.
During initialization, QuadTree decomposition is done and result are store in self.blocks as a list containing [x,y,height, width,Std].
To visualize the results, use plot().
-
plot(edgecolor='red', facecolor='none', linewidth=1)¶ This function is used to generate a graphical representation of the QuadTree decomposition.
Parameters: - edgecolor (string) – color of the rectangles, default is red
- facecolor (string) – color used for rectangles fills. Default is none.
- linewidth – width in px of the rectangles’ borders. Default is 1.
Returns: plot with image and leaves of the quadTree Decomposition
-
quadTreeDecomposition(img, x, y, minStd, minSize)¶ This function evaluate the mean and std of an image, and decides Whether to perform or not other 2 splits of the leave.
Parameters: - img (numpy.ndarray) – img to analyze
- x (int) – x offset of the leaves to analyze
- Y (int) – Y offset of the leaves to analyze
MinStd: Std threshold for subsequent splitting
MinSize: Size threshold for subsequent splitting, in pixel
-