Package Msmoothgradient

Return gradient of input image with smoothing orthogonal to gradient direction. More...

Variables
tuple	par = sf.Par()
tuple	input = sf.Input()
tuple	output1 = sf.Output(par.string("dx_outputfilename"))
tuple	output2 = sf.Output(par.string("dy_outputfilename"))
tuple	n1 = input.int("n1")
tuple	n2 = input.int("n2")
tuple	ni = input.size(2)
tuple	inputimage = np.zeros((n1,n2),'f')
tuple	input_padded = Sadiabatic(inputimage)
tuple	I = array(range(1,n1+1))
tuple	J = array(range(1,n2+1))
tuple	Imx = (3*input_padded[I+1][:,J-1]+10*input_padded[I+1][:,J]+3*input_padded[I+1][:,J+1]-3*input_padded[I-1][:,J-1]-10*input_padded[I-1][:,J]-3*input_padded[I-1][:,J+1])
tuple	Imy = (3*input_padded[I-1][:,J+1]+10*input_padded[I][:,J+1]+3*input_padded[I+1][:,J+1]-3*input_padded[I-1][:,J-1]-10*input_padded[I][:,J-1]-3*input_padded[I+1][:,J-1])

---

Detailed Description

Return gradient of input image with smoothing orthogonal to gradient direction.

 import rsf as sf
 import numpy as np

Import RSF and NumPy packages as they will be used later in the code

 from Sadiabatic import *

Sadiabatic is a package containing a function of the same name used for padding the boundaries of the image with repetition (Neumann boundary conditions)

Get the RSF parameters for this function

 par = sf.Par()

Get the name of the file for which we want to calculate the gradients from the RSF input. As well, output1 is the name of the RSF file to which we write the gradient values in the x direction, and output2 is the RSF filename for writing the gradient values in the y direction.

 input  = sf.Input()
 output1 = sf.Input(par.string("dx_outputfilename"))
 output2 = sf.Input(par.string("dy_outputfilename"))

n1 is the width of the image and n2 is the height of the image

 n1 = input.int("n1")
 n2 = input.int("n2")

Make sure the input is 2-D. Otherwise, the gradients cannot be calculated with this code.

 ni = input.size(2)
 assert ni == 1,"sfsmoothgradient needs 2D input"

Read in the input image

 inputimage = np.zeros((n1,n2),'f')
 input.read(inputimage)

Pad the input image with repetition so that gradients can be calculated on the boundaries.

 input_padded = Sadiabatic(inputimage)

Used for indexing of image for computation of gradient

 I = array(range(1,n1+1))
 J = array(range(1,n2+1))

Calculate gradient in x direction

 Imx = (3*input_padded[I+1][:,J-1]+10*input_padded[I+1][:,J]+3*input_padded[I+1][:,J+1]-3*input_padded[I-1][:,J-1]-10*input_padded[I-1][:,J]-3*input_padded[I-1][:,J+1])/32

Calculate gradient in y direction

 Imy = (3*input_padded[I-1][:,J+1]+10*input_padded[I][:,J+1]+3*input_padded[I+1][:,J+1]-3*input_padded[I-1][:,J-1]-10*input_padded[I][:,J-1]-3*input_padded[I+1][:,J-1])/32

Write directional gradients (in x and y directions) to appropriate output files

 output1.put("n1",n1)
 output1.put("n2",n2)
 output1.write(Imx)
 
 output2.put("n1",n1)
 output2.put("n2",n2)
 output2.write(Imy)

---

Variable Documentation

tuple I = array(range(1,n1+1))

Definition at line 129 of file Msmoothgradient.py.

tuple Imx = (3*input_padded[I+1][:,J-1]+10*input_padded[I+1][:,J]+3*input_padded[I+1][:,J+1]-3*input_padded[I-1][:,J-1]-10*input_padded[I-1][:,J]-3*input_padded[I-1][:,J+1])

Definition at line 133 of file Msmoothgradient.py.

tuple Imy = (3*input_padded[I-1][:,J+1]+10*input_padded[I][:,J+1]+3*input_padded[I+1][:,J+1]-3*input_padded[I-1][:,J-1]-10*input_padded[I][:,J-1]-3*input_padded[I+1][:,J-1])

Definition at line 136 of file Msmoothgradient.py.

tuple input = sf.Input()

Definition at line 112 of file Msmoothgradient.py.

tuple input_padded = Sadiabatic(inputimage)

Definition at line 126 of file Msmoothgradient.py.

tuple inputimage = np.zeros((n1,n2),'f')

Definition at line 124 of file Msmoothgradient.py.

tuple J = array(range(1,n2+1))

Definition at line 130 of file Msmoothgradient.py.

tuple n1 = input.int("n1")

Definition at line 119 of file Msmoothgradient.py.

tuple n2 = input.int("n2")

Definition at line 120 of file Msmoothgradient.py.

tuple ni = input.size(2)

Definition at line 121 of file Msmoothgradient.py.

tuple output1 = sf.Output(par.string("dx_outputfilename"))

Definition at line 116 of file Msmoothgradient.py.

tuple output2 = sf.Output(par.string("dy_outputfilename"))

Definition at line 117 of file Msmoothgradient.py.

tuple par = sf.Par()

Definition at line 110 of file Msmoothgradient.py.