skimage_segmenter.py 4.37 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118
#!/usr/bin/python
# -*- coding: utf-8 -*-
#
"""
    Abstract class for segmenters implemented in skimage.segmentation.
    
    Name: skimage_segmenter.py
    Author: Alessandro dos Santos Ferreira ( santosferreira.alessandro@gmail.com )
"""

import cv2
import numpy as np
from skimage.segmentation import mark_boundaries
from skimage.util import img_as_float, img_as_ubyte

from util.config import Config
from util.utils import TimeUtils
from util.x11_colors import X11Colors

from abc import ABCMeta, abstractmethod

class SkimageSegmenter(object):
    
    __metaclass__ = ABCMeta
    
    def __init__(self, border_color = 'Yellow', border_outline = 'No'):
        self.border_color = Config("Border Color", border_color, 'color')
        self.border_outline = Config("Border Outline", border_outline, str)
        
        self._segments = None
        self._original_image = None
        
    
    def get_segment_skimage(self, px = 0, py = 0, idx_segment = None):
        if self._segments is None:
            return None, 0, -1, 0
        
        start_time = TimeUtils.get_time()
        
        if idx_segment is None:
            idx_segment = self._segments[py, px]
        
        mask_segment = np.zeros(self._original_image.shape[:2], dtype="uint8")
        mask_segment[self._segments == idx_segment] = 255
        size_segment = mask_segment[self._segments == idx_segment].size

        segment = self._original_image.copy()
        segment = cv2.bitwise_and(segment, segment, mask=mask_segment)

        contours, _  = cv2.findContours(mask_segment,cv2.RETR_LIST,cv2.CHAIN_APPROX_SIMPLE)[-2:]

        m = -1
        max_contour = None
        for cnt in contours:
            if (len(cnt) > m):
                m = len(cnt)
                max_contour = cnt

        x,y,w,h = cv2.boundingRect(max_contour)
        segment = segment[y:y+h, x:x+w]
        
        end_time = TimeUtils.get_time()
        
        return segment, size_segment, idx_segment, (end_time - start_time)
        
        
    def paint_segment_skimage(self, image, color, px = 0, py = 0, idx_segment = None, border = True, clear = False):
        if self._segments is None:
            return image, 0
            
        start_time = TimeUtils.get_time()
        
        if idx_segment is None:
            idx_segment = self._segments[py, px]
        height, width, channels = self._original_image.shape
        
        mask_segment = np.zeros(self._original_image.shape[:2], dtype="uint8")
        mask_segment[self._segments == idx_segment] = 255
        mask_inv = cv2.bitwise_not(mask_segment)
            
        class_color = np.zeros((height,width,3), np.uint8)
        class_color[:, :] = X11Colors.get_color(color)
        if clear == False:
            colored_image = cv2.addWeighted(self._original_image, 0.7, class_color, 0.3, 0)
        else:
            colored_image = self._original_image
        colored_image = cv2.bitwise_and(colored_image, colored_image, mask=mask_segment)
        
        new_image = cv2.bitwise_and(image, image, mask=mask_inv)
        mask_segment[:] = 255
        new_image = cv2.bitwise_or(new_image, colored_image, mask=mask_segment)

        if border == True:
            color = X11Colors.get_color_zero_one(self.border_color.get_cast_val())
            outline_color = color if self.border_outline.value == 'Yes' else None
            
            new_image = img_as_ubyte( mark_boundaries(img_as_float(new_image), self._segments.astype(np.int8), color=color, outline_color=outline_color) ) 
            
        end_time = TimeUtils.get_time()
        
        return new_image, (end_time - start_time)


    def run_skimage(self, image, method, **kwargs):
        self._original_image = image
        
        start_time = TimeUtils.get_time()
        self._segments = method(img_as_float(image), **kwargs)
        end_time = TimeUtils.get_time()
        
        color = X11Colors.get_color_zero_one(self.border_color.get_cast_val())
        outline_color = color if self.border_outline.value == 'Yes' else None

        #  Ignore UserWarning: Possible precision loss when converting from float64 to uint8
        #  because the returned image is used just for visualization
        #  The original image, without loss, is stored in self._original_image
        return img_as_ubyte( mark_boundaries(image, self._segments.astype(np.int8), color=color, outline_color=outline_color) ), (end_time - start_time)