Added abd replaced some filters, added docstrings.
This commit is contained in:
152
src/filters.py
152
src/filters.py
@ -1,17 +1,18 @@
|
||||
"""! @file filters.py
|
||||
@brief Filter library for the application
|
||||
@author xlanro00
|
||||
@brief Filter library for the application
|
||||
@author xlanro00
|
||||
"""
|
||||
|
||||
import numpy as np
|
||||
#import matplotlib.pyplot as plt
|
||||
import cv2 as cv
|
||||
from skimage import filters as skiflt
|
||||
from skimage import restoration as skirest
|
||||
|
||||
|
||||
# Parent class for all the filters
|
||||
class filter:
|
||||
'''
|
||||
Parent class for all the filters
|
||||
''' Parent class for all the filters.
|
||||
'''
|
||||
|
||||
def __init__(self, img):
|
||||
@ -19,8 +20,9 @@ class filter:
|
||||
|
||||
|
||||
class convolve(filter):
|
||||
''' Convolve using custom kernel,
|
||||
if no kernel is given, use default 3x3 kernel for averaging
|
||||
''' Convolve with custom kernel using opencv.
|
||||
If no kernel is given, use default 3x3 kernel for averaging.
|
||||
Possibly useful for custom filters.
|
||||
'''
|
||||
|
||||
def __init__(self, img):
|
||||
@ -31,15 +33,20 @@ class convolve(filter):
|
||||
kernel = np.array(params["kernel"]) if params["kernel"] else np.ones(
|
||||
(ksize, ksize), np.float32) / np.sqrt(ksize)
|
||||
|
||||
#print("with params: " + " ksize: " + str(ksize) + " kernel: \n" + str(kernel))
|
||||
print("with params: ksize: " +
|
||||
str(ksize) + " kernel: \n" + str(kernel))
|
||||
self.img = cv.filter2D(self.img, -1, kernel)
|
||||
|
||||
|
||||
class blur(filter):
|
||||
''' Blur filter from OpenCV.
|
||||
Performs averaging of the image.
|
||||
'''
|
||||
def __init__(self, img):
|
||||
super().__init__(img)
|
||||
|
||||
def apply(self, params):
|
||||
# TODO remove try-except
|
||||
if(params["anchor"]):
|
||||
try:
|
||||
anchor = tuple(map(int, params["anchor"].split(',')))
|
||||
@ -49,11 +56,14 @@ class blur(filter):
|
||||
anchor = (-1, -1)
|
||||
ksize = int(params["ksize"]) if params["ksize"] else 3
|
||||
|
||||
#print("with params: " + " ksize: " + str(ksize) + " anchor: " + str(anchor))
|
||||
print("with params: ksize: " +
|
||||
str(ksize) + " anchor: " + str(anchor))
|
||||
self.img = cv.blur(self.img, ksize=(ksize, ksize), anchor=anchor)
|
||||
|
||||
|
||||
class gaussian(filter):
|
||||
''' Gaussian blur filter from OpenCV.
|
||||
'''
|
||||
def __init__(self, img):
|
||||
super().__init__(img)
|
||||
|
||||
@ -62,22 +72,27 @@ class gaussian(filter):
|
||||
sigmaX = float(params["sigmaX"]) if params["sigmaX"] else 0
|
||||
sigmaY = float(params["sigmaY"]) if params["sigmaY"] else 0
|
||||
|
||||
#print("with params: " + " ksize: " + str(ksize) + " sigmaX: " + str(sigmaX) + " sigmaY: " + str(sigmaY))
|
||||
print("with params: ksize: " + str(ksize) +
|
||||
" sigmaX: " + str(sigmaX) + " sigmaY: " + str(sigmaY))
|
||||
self.img = cv.GaussianBlur(self.img, (ksize, ksize), sigmaX, sigmaY)
|
||||
|
||||
|
||||
class median(filter):
|
||||
''' Median blur filter from OpenCV.
|
||||
'''
|
||||
def __init__(self, img):
|
||||
super().__init__(img)
|
||||
|
||||
def apply(self, params):
|
||||
ksize = int(params["ksize"]) if params["ksize"] else 3
|
||||
|
||||
#print("with params: " + " ksize: " + str(ksize))
|
||||
self.img = cv.medianBlur(np.uint8(self.img), ksize)
|
||||
print("with params: ksize: " + str(ksize))
|
||||
self.img = cv.medianBlur(np.float32(self.img), ksize)
|
||||
|
||||
|
||||
class bilateral(filter):
|
||||
''' Bilateral filter from OpenCV.
|
||||
'''
|
||||
def __init__(self, img):
|
||||
super().__init__(img)
|
||||
|
||||
@ -87,27 +102,87 @@ class bilateral(filter):
|
||||
sigmaColor = int(params["sigmaColor"]) if params["sigmaColor"] else 75
|
||||
sigmaSpace = int(params["sigmaSpace"]) if params["sigmaSpace"] else 75
|
||||
|
||||
#print("with params: " + " d: " + str(d) + " sigmaColor: " + str(sigmaColor) + " sigmaSpace: " + str(sigmaSpace))
|
||||
print("with params: d: " + str(d) + " sigmaColor: " +
|
||||
str(sigmaColor) + " sigmaSpace: " + str(sigmaSpace))
|
||||
self.img = cv.bilateralFilter(self.img, d, sigmaColor, sigmaSpace)
|
||||
|
||||
|
||||
class denoise(filter):
|
||||
# TODO possibly not necessary
|
||||
def __init__(self, img):
|
||||
super().__init__(img)
|
||||
|
||||
def apply(self, params):
|
||||
h = int(params["h"]) if params["h"] else 20
|
||||
h = int(params["h"]) if params["h"] else 10
|
||||
tWS = int(params["templateWindowSize"]
|
||||
) if params["templateWindowSize"] else 7
|
||||
sWS = int(params["searchWindowSize"]
|
||||
) if params["searchWindowSize"] else 21
|
||||
|
||||
#print("with params: " + " h: " + str(h) + " tWS: " + str(tWS) + " sWS: " + str(sWS))
|
||||
print("with params: h: " + str(h) +
|
||||
" tWS: " + str(tWS) + " sWS: " + str(sWS))
|
||||
self.img = np.uint8(self.img)
|
||||
self.img = cv.fastNlMeansDenoising(self.img, h, tWS, sWS)
|
||||
self.img = cv.fastNlMeansDenoising(
|
||||
self.img, h, tWS, sWS)
|
||||
|
||||
|
||||
class denoise_bilateral(filter):
|
||||
''' Scikit image denoise_bilateral filter.
|
||||
|
||||
Performs bilateral denoising technique on the image.
|
||||
Averages pixels based on their distance and color similarity.
|
||||
Preserves edges while removing unwanted noise.
|
||||
'''
|
||||
|
||||
def __init__(self, img):
|
||||
super().__init__(img)
|
||||
|
||||
def apply(self, params):
|
||||
sigmaColor = float(params["sigmaColor"]
|
||||
) if params["sigmaColor"] else 0.1
|
||||
sigmaSpace = float(params["sigmaSpace"]
|
||||
) if params["sigmaSpace"] else 15.0
|
||||
channelAxis = int(params["channelAxis"]
|
||||
) if params["channelAxis"] else None
|
||||
iterations = int(params["iterations"]) if params["iterations"] else 1
|
||||
|
||||
print("with params: sigma_color: " + str(sigmaColor) +
|
||||
" sigma_spatial: " + str(sigmaSpace) + " channel_axis: " +
|
||||
str(channelAxis) + " iterations: " + str(iterations))
|
||||
|
||||
for i in range(iterations):
|
||||
self.img = skirest.denoise_bilateral(
|
||||
self.img, sigma_color=sigmaColor,
|
||||
sigma_spatial=sigmaSpace, channel_axis=channelAxis)
|
||||
|
||||
|
||||
class denoise_tv_chambolle(filter):
|
||||
''' Scikit image denoise_tv_chambolle filter from scikit-image.
|
||||
|
||||
Performs total variation denoising technique on the image.
|
||||
This filter removes fine detail, but preserves edges.
|
||||
'''
|
||||
|
||||
def __init__(self, img):
|
||||
super().__init__(img)
|
||||
|
||||
def apply(self, params):
|
||||
weight = float(params["weight"]) if params["weight"] else 0.1
|
||||
channelAxis = int(params["channelAxis"]
|
||||
) if params["channelAxis"] else None
|
||||
iterations = int(params["iterations"]) if params["iterations"] else 1
|
||||
|
||||
print("with params: weight: " + str(weight) +
|
||||
" channel_axis: " + str(channelAxis) + " iterations: " + str(iterations))
|
||||
for i in range(iterations):
|
||||
self.img = skirest.denoise_tv_chambolle(
|
||||
self.img, weight=weight, channel_axis=channelAxis)
|
||||
|
||||
|
||||
class sharpen(filter):
|
||||
''' Convolution with a sharpening kernel using opencv.
|
||||
'''
|
||||
# TODO possibly unnecessary, because unsharp masking is working better
|
||||
def __init__(self, img):
|
||||
super().__init__(img)
|
||||
|
||||
@ -115,13 +190,13 @@ class sharpen(filter):
|
||||
kernel = np.matrix(params["kernel"]) if params["kernel"] else np.array(
|
||||
[[0, -1, 0], [-1, 5, -1], [0, -1, 0]])
|
||||
|
||||
#print("with params: " + " kernel: \n" + str(kernel))
|
||||
print("with params: kernel: \n" + str(kernel))
|
||||
self.img = cv.filter2D(self.img, ddepth=-1, kernel=kernel)
|
||||
|
||||
|
||||
class unsharp_mask(filter):
|
||||
''' Unsharp mask filter.
|
||||
|
||||
''' Unsharp mask filter from opencv.
|
||||
|
||||
First blur the image a little bit, then calculate Laplacian of the image to get the edges.
|
||||
Scale the Laplacian and subtract it from the original image.
|
||||
'''
|
||||
@ -132,13 +207,43 @@ class unsharp_mask(filter):
|
||||
def apply(self, params):
|
||||
strength = float(params["strength"]) if params["strength"] else 1.0
|
||||
ksize = int(params["ksize"]) if params["ksize"] else 3
|
||||
blurred = cv.medianBlur(np.uint8(self.img), ksize)
|
||||
|
||||
blurred = cv.medianBlur(self.img, ksize)
|
||||
lap = cv.Laplacian(blurred, cv.CV_32F)
|
||||
|
||||
print("with params: strength: " +
|
||||
str(strength) + " ksize: " + str(ksize))
|
||||
self.img = blurred - strength*lap
|
||||
|
||||
|
||||
class unsharp_mask_scikit(filter):
|
||||
''' Unsharp mask filter from scikit.
|
||||
|
||||
Apply blurring using gaussian filter, then subtract the blurred image from the original image.
|
||||
Radius parameter is the sigma parameter of the gaussian filter.
|
||||
Amount parameter regulates the strength of the unsharp mask.
|
||||
Better results than using opencv module.
|
||||
'''
|
||||
|
||||
def __init__(self, img):
|
||||
super().__init__(img)
|
||||
|
||||
def apply(self, params):
|
||||
radius = int(params["radius"]) if params["radius"] else 3
|
||||
amount = float(params["amount"]) if params["amount"] else 1
|
||||
channelAxis = int(params["channelAxis"]
|
||||
) if params["channelAxis"] else None
|
||||
|
||||
#self.img = cv.cvtColor(self.img, cv.COLOR_GRAY2RGB)
|
||||
print("with params: radius: " +
|
||||
str(radius) + " amount: " + str(amount))
|
||||
self.img = skiflt.unsharp_mask(
|
||||
self.img, radius=radius, amount=amount, channel_axis=channelAxis)
|
||||
#self.img = cv.cvtColor(self.img, cv.COLOR_RGB2GRAY)
|
||||
|
||||
|
||||
class morph(filter):
|
||||
''' General morphological operations.
|
||||
''' General morphological operations from OpenCV.
|
||||
|
||||
Can be used with MORPH_OPEN, MORPH_CLOSE, MORPH_DILATE, MORPH_ERODE and more as 'op'.
|
||||
'''
|
||||
@ -152,6 +257,7 @@ class morph(filter):
|
||||
iterations = int(params["iterations"]) if params["iterations"] else 1
|
||||
op = getattr(cv, params["op"]) if params["op"] else cv.MORPH_OPEN
|
||||
if(params["anchor"]):
|
||||
# TODO remove try-except
|
||||
try:
|
||||
anchor = tuple(map(int, params["anchor"].split(',')))
|
||||
except AttributeError:
|
||||
@ -159,6 +265,8 @@ class morph(filter):
|
||||
else:
|
||||
anchor = (-1, -1)
|
||||
|
||||
#print("with params: " + " kernel: \n" + str(kernel) + " anchor: " + str(anchor) + " iterations: " + str(iterations) + " op: " + str(op))
|
||||
print("with params: kernel: \n" + str(kernel) + " anchor: " +
|
||||
str(anchor) + " iterations: " + str(iterations) + " op: " + str(op))
|
||||
self.img = cv.morphologyEx(
|
||||
self.img, op=op, kernel=kernel, anchor=anchor, iterations=iterations)
|
||||
np.uint8(self.img), op=op, kernel=kernel,
|
||||
anchor=anchor, iterations=iterations)
|
||||
|
81
src/main.py
81
src/main.py
@ -1,13 +1,12 @@
|
||||
"""! @file main.py
|
||||
@brief Main file for the application
|
||||
@author xlanro00
|
||||
@brief Main file for the application
|
||||
@author xlanro00
|
||||
"""
|
||||
|
||||
# Import basic libraries
|
||||
import argparse as ap
|
||||
import sys
|
||||
import json
|
||||
#from datetime import datetime
|
||||
|
||||
# Libraries for image processing
|
||||
import numpy as np
|
||||
@ -33,7 +32,6 @@ class apply_filters:
|
||||
self.config_file = self.args.config[0]
|
||||
self.preset_name = self.args.config[1]
|
||||
self.config = json.load(open(self.config_file))
|
||||
print("Config loaded")
|
||||
self.parse_conf()
|
||||
|
||||
# If no config file given, expect filters in command line
|
||||
@ -63,11 +61,12 @@ class apply_filters:
|
||||
|
||||
def run(self):
|
||||
# read as numpy.array
|
||||
self.img = cv.imread(self.input_file, cv.IMREAD_GRAYSCALE)
|
||||
self.img = cv.imread(
|
||||
self.input_file, cv.IMREAD_GRAYSCALE).astype(np.uint8)
|
||||
|
||||
self.width = self.img.shape[1]
|
||||
self.height = self.img.shape[0]
|
||||
print(self.width, self.height)
|
||||
self.print_size(self.img.shape)
|
||||
|
||||
fig = plt.figure(figsize=(self.width, self.height),
|
||||
frameon=False, dpi=self.dpi / 100) # dpi is in cm
|
||||
@ -86,21 +85,27 @@ class apply_filters:
|
||||
if self.args.stl:
|
||||
self.make_lithophane()
|
||||
|
||||
|
||||
def parse_params(self, params):
|
||||
''' Parse parameters of filters.
|
||||
Set to None if not given.
|
||||
They are later set in the filter method.
|
||||
'''
|
||||
|
||||
possible_params = {"h", "searchWindowSize", "templateWindowSize",
|
||||
"ksize", "kernel", "sigmaX", "sigmaY",
|
||||
"sigmaColor", "sigmaSpace", "d", "anchor", "iterations",
|
||||
"op", "strength"}
|
||||
"sigmaColor", "sigmaSpace", "d", "anchor", "iterations",
|
||||
"op", "strength", "amount", "radius", "weight", "channelAxis"}
|
||||
for key in possible_params:
|
||||
try:
|
||||
params[key] = params[key]
|
||||
except KeyError:
|
||||
if params.get(key) is None:
|
||||
params[key] = None
|
||||
else:
|
||||
params[key] = params[key]
|
||||
|
||||
def parse_conf(self):
|
||||
# Parse configuration file if given.
|
||||
try:
|
||||
''' Parse configuration file if one was given and store filters with their parameters
|
||||
'''
|
||||
|
||||
if self.preset_name in self.config:
|
||||
filter_array = self.config[self.preset_name]
|
||||
for i, filter in enumerate(range(len(filter_array)), start=1):
|
||||
self.filters.append(filter_array[filter]["name"])
|
||||
@ -109,13 +114,15 @@ class apply_filters:
|
||||
if attribute != "name":
|
||||
self.params[i][attribute] = value
|
||||
self.parse_params(self.params[i])
|
||||
|
||||
except(KeyError):
|
||||
print("Loaded preset: " + self.preset_name +
|
||||
" from file: " + self.config_file)
|
||||
else:
|
||||
print("Preset not found", file=sys.stderr)
|
||||
|
||||
def parse_arguments(self):
|
||||
''' Parse arguments from command line
|
||||
'''
|
||||
|
||||
# Parse arguments
|
||||
parser = ap.ArgumentParser(prog='main.py',
|
||||
description='Program for processing a 2D image into 3D fingerprint.',
|
||||
usage='%(prog)s [-h] [-m | --mirror | --no-mirror] input_file output_file dpi ([-c config_file preset | --config config_file preset] | [filters ...])')
|
||||
@ -127,10 +134,11 @@ class apply_filters:
|
||||
help="output file location")
|
||||
parser.add_argument("dpi", type=int, help="scanner dpi")
|
||||
|
||||
# boolean switch
|
||||
# boolean switch argument
|
||||
parser.add_argument('-m', "--mirror", help="mirror input image",
|
||||
type=bool, action=ap.BooleanOptionalAction)
|
||||
|
||||
# another boolean switch argument
|
||||
parser.add_argument('-s', '--stl', help="make stl model from processed image",
|
||||
type=bool, action=ap.BooleanOptionalAction)
|
||||
|
||||
@ -149,7 +157,7 @@ class apply_filters:
|
||||
''' Selects filter method of filters library.
|
||||
'''
|
||||
|
||||
print("Applying " + filter_name + " filter", file=sys.stderr)
|
||||
print("Applying " + filter_name + " filter ", end='')
|
||||
return getattr(flt, filter_name)
|
||||
|
||||
def resize_image(self):
|
||||
@ -163,16 +171,16 @@ class apply_filters:
|
||||
should be used only if we want a positive model
|
||||
'''
|
||||
|
||||
#TODO make this automatic for positive STL
|
||||
# TODO make this automatic for positive STL
|
||||
print("Mirroring image", file=sys.stderr)
|
||||
self.img = cv.flip(self.img, 1) # 1 for vertical mirror
|
||||
|
||||
def apply_filter(self):
|
||||
''' Apply filters to image.
|
||||
|
||||
|
||||
Applies the filters one by one, if no filters were given, just save original image output.
|
||||
'''
|
||||
|
||||
|
||||
if len(self.filters) == 0:
|
||||
# No filter given, just save the image
|
||||
pass
|
||||
@ -180,11 +188,13 @@ class apply_filters:
|
||||
# Apply all filters
|
||||
for i, filter_name in enumerate(self.filters):
|
||||
filter = self.filter_factory(filter_name)
|
||||
# print(self.img.dtype)
|
||||
filter.apply(self, self.params[i+1])
|
||||
# print(self.img.dtype)
|
||||
|
||||
def print_size(self, size):
|
||||
print("Width: " + str(size[0]), file=sys.stderr)
|
||||
print("Height: " + str(size[1]), file=sys.stderr)
|
||||
print("Height: " + str(size[0]), file=sys.stderr)
|
||||
print("Width: " + str(size[1]), file=sys.stderr)
|
||||
|
||||
def save_image(self, fig, ax):
|
||||
''' Save processed image.
|
||||
@ -196,7 +206,6 @@ class apply_filters:
|
||||
fig.savefig(fname=self.output_file)
|
||||
|
||||
def make_lithophane(self):
|
||||
pass
|
||||
'''After processing image, make a lithophane from it.
|
||||
'''
|
||||
|
||||
@ -208,15 +217,21 @@ class apply_filters:
|
||||
self.save_model()
|
||||
|
||||
def make_meshgrid(self):
|
||||
|
||||
''' Create numpy meshgrid.
|
||||
Modify image values to get more usable depth values.
|
||||
Add zero padding to image to make sides of the plate.
|
||||
'''
|
||||
# Modify image to make it more suitable depth
|
||||
# values1 = (1 + (1 - self.img/255)/6) * 255/10 # this works
|
||||
# values2 = (1 - (1 - self.img/255)/6) * 255/10 # TODO: i dont know how to make white surrounding be extruded
|
||||
# values2 = (1 - (1 - self.img/255)/6) * 255/10 #
|
||||
# TODO: i dont know how to make white surrounding be extruded
|
||||
|
||||
values1better = 28.05 - 0.01*self.img
|
||||
#values2better = 22.95 - 0.01*self.img
|
||||
# (np.around(values2[::300],3))
|
||||
|
||||
# Add zero padding to image to make sides of the plate
|
||||
# Add zero padding to image
|
||||
# TODO this better be done in the next function to keep dimensions intact
|
||||
self.height = self.img.shape[0] + 2
|
||||
self.width = self.img.shape[1] + 2
|
||||
self.img = np.zeros([self.height, self.width])
|
||||
@ -228,6 +243,11 @@ class apply_filters:
|
||||
self.meshgrid = np.meshgrid(verticesX, verticesY)
|
||||
|
||||
def make_mesh(self):
|
||||
''' Create mesh from image.
|
||||
Create vertices from meshgrid, add depth values from image.
|
||||
Create faces from vertices.
|
||||
'''
|
||||
|
||||
# Convert meshgrid and image matrix to array of 3D points
|
||||
vertice_arr = np.vstack(list(map(np.ravel, self.meshgrid))).T
|
||||
z = (self.img / 10).reshape(-1, 1)
|
||||
@ -278,7 +298,10 @@ class apply_filters:
|
||||
self.model.vectors[i][j] = vertices[face[j], :]
|
||||
|
||||
def save_model(self):
|
||||
print("Saving stl model", file=sys.stderr)
|
||||
''' Save final model to stl file.
|
||||
'''
|
||||
|
||||
print("Saving lithophane to stl file", file=sys.stderr)
|
||||
self.model.save('res/test.stl')
|
||||
|
||||
|
||||
|
Reference in New Issue
Block a user