一、研究背景
在上一篇slic超像素分割的文章中，提到了需要对分割后的超像素进行特征提取，依旧为简化复现论文，论文在文末附上，在这里我整理以下特征提取过程中的代码以及心得，方便后期回溯复盘。本文的特征提取均在slic算法基础上进行研究。
二、特征提取
在针对一个乳腺超声图片分割出来的超像素进行分类的过程中，我们需要提取到一些分类的依据，比如颜色特征，纹理特征等等。并且需要将每一个超像素打上是否为恶性肿瘤的标签，以便将当前超像素所拥有的特征进行划分。
1、颜色特征
在slic算法中，每一个像素都有自己的l，a，b值，一个超像素如果只提取l，a，b值将会得到很大的一组值，我们为了方便起见只提取一个超像素中所有像素l，a，b的平均值与方差。代码如下
def average(self):#数据平均值
for cluster in self.clusters: #每一个超像素
l = a = b = number =0
for p in cluster.pixels: #超像素中的每一个像素
l += self.data[p[0]][p[1]][0]
a += self.data[p[0]][p[1]][1]
b += self.data[p[0]][p[1]][2]
number += 1 #像素总个数
strin=str([l/number,a/number,b/number]) #均值
with open(f'C:/Users/Administrator/Desktop/data/{self.n}/benign{self.n}_average.txt', 'a') as f: #保存数据
f.write('\n' + strin)
def fangcha(self):#数据方差
for cluster in self.clusters:
l = a = b = []
for p in cluster.pixels:
l.append(self.data[p[0]][p[1]][0])
a.append(self.data[p[0]][p[1]][1])
b.append(self.data[p[0]][p[1]][2])
stri=str([np.var(l),np.var(a),np.var(b)]) #方差
with open(f'C:/Users/Administrator/Desktop/data/{self.n}/benign{self.n}_variance.txt', 'a') as f:
f.write('\n' + stri)2、纹理特征
纹理特征的提取主要使用了灰度共生矩阵
我参考了部分python代码，结合slic代码写了一下关于逐像素进行灰度共生矩阵的生成
注：灰度共生矩阵的生成与输入图片是否规则没有关系！
def getGlcm(self,d_y,d_x):#灰度共生矩阵
#将一整张图片转化为灰度图
glcm_img = cv2.imread(f"C:/Users/Administrator/Desktop/SRP/Dataset_BUSI_with_GT/benign/benign ({self.n}).png", 0)
for cluster in self.clusters: #每个超像素块
max_gray_level = number=0
ret = [[0.0 for i in range(gray_level)] for j in range(gray_level)]
for p in cluster.pixels:
number += 1
if glcm_img[p[0]][p[1]] > max_gray_level:
max_gray_level = glcm_img[p[0]][p[1]]
max_gray_level = max_gray_level + 1 #得到此超像素块中最大灰度级
for pix in cluster.pixels:
if max_gray_level > gray_level: #若是最大灰度级大于设定灰度级，则将其调整为设定灰度级大小
glcm_img[pix[0]][pix[1]] = glcm_img[pix[0]][pix[1]] * gray_level / max_gray_level
for pixe in cluster.pixels: #再次逐像素遍历
h_1 = pixe[0] + d_y #设定边界
w_1 = pixe[1] + d_x
if h_1 > glcm_img.shape[0] - 1:
h_1 = glcm_img.shape[0] - 1
if w_1 > glcm_img.shape[1] - 1:
w_1 = glcm_img.shape[1] - 1
rows = glcm_img[pixe[0]][pixe[1]]
cols = glcm_img[h_1][w_1]
if rows >= 16:
rows = 15
if cols >= 16:
cols = 15
ret[rows][cols] += 1.0
for i in range(gray_level):
for j in range(gray_level):
ret[i][j] /= float(number) #得到灰度生成矩阵
asm, con, eng, idm = feature_computer(ret)
string=str([asm, con, eng, idm])
with open(f'C:/Users/Administrator/Desktop/data/{self.n}/benign{self.n}_d_y={d_y}_d_x={d_x}.txt', 'a') as f:
f.write('\n'+ string)三、完整代码
import math
import cv2
from skimage import io, color
import numpy as np
from tqdm import trange
gray_level=16
def feature_computer(p):#GLCM的特征提取
# con:对比度反应了图像的清晰度和纹理的沟纹深浅。纹理越清晰反差越大对比度也就越大。
# eng:熵（Entropy, ENT）度量了图像包含信息量的随机性，表现了图像的复杂程度。当共生矩阵中所有值均相等或者像素值表现出最大的随机性时，熵最大。
# agm:角二阶矩（能量），图像灰度分布均匀程度和纹理粗细的度量。当图像纹理均一规则时，能量值较大；反之灰度共生矩阵的元素值相近，能量值较小。
# idm:反差分矩阵又称逆方差，反映了纹理的清晰程度和规则程度，纹理清晰、规律性较强、易于描述的，值较大。
Con = 0.0
Eng = 0.0
Asm = 0.0
Idm = 0.0
for i in range(gray_level):
for j in range(gray_level):
Con += (i - j) * (i - j) * p[i][j]
Asm += p[i][j] * p[i][j]
Idm += p[i][j] / (1 + (i - j) * (i - j))
if p[i][j] > 0.0:
Eng += p[i][j] * math.log(p[i][j])
return Asm, Con, -Eng, Idm
class Cluster(object):
cluster_index = 1
def __init__(self, h, w, l=0, a=0, b=0): #初始化
self.update(h, w, l, a, b)
self.pixels = []
self.no = self.cluster_index
Cluster.cluster_index += 1
def update(self, h, w, l, a, b):
self.h = h
self.w = w
self.l = l
self.a = a
self.b = b
def __str__(self):
return "{},{}:{} {} {} ".format(self.h, self.w, self.l, self.a, self.b)
def __repr__(self):
return self.__str__()
class SLICProcessor(object):
@staticmethod
def open_image(path):#将rgb图片转为lab图片
rgb = io.imread(path)
lab_arr = color.rgb2lab(rgb)
return lab_arr
@staticmethod
def save_lab_image(path, lab_arr):
rgb_arr = color.lab2rgb(lab_arr)
io.imsave(path, rgb_arr)
def make_cluster(self, h, w):
h = int(h)
w = int(w)
return Cluster(h, w,
self.data[h][w][0],
self.data[h][w][1],
self.data[h][w][2])
def __init__(self, filename, K, M,n):
self.K = K
self.M = M
self.n = n #第几张图片
self.data = self.open_image(filename)
self.image_height = self.data.shape[0]
self.image_width = self.data.shape[1]
self.N = self.image_height * self.image_width
self.S = int(math.sqrt(self.N / self.K))
self.clusters = []
self.label = {}
self.dis = np.full((self.image_height, self.image_width), np.inf)
def init_clusters(self):
h = self.S / 2
w = self.S / 2
while h < self.image_height:
while w < self.image_width:
self.clusters.append(self.make_cluster(h, w))
w += self.S
w = self.S / 2
h += self.S
def get_gradient(self, h, w):
if w + 1 >= self.image_width:
w = self.image_width - 2
if h + 1 >= self.image_height:
h = self.image_height - 2
gradient = self.data[h + 1][w + 1][0] - self.data[h][w][0] + \
self.data[h + 1][w + 1][1] - self.data[h][w][1] + \
self.data[h + 1][w + 1][2] - self.data[h][w][2]
return gradient
def move_clusters(self):#确定聚类中心点
for cluster in self.clusters:
cluster_gradient = self.get_gradient(cluster.h, cluster.w)
for dh in range(-1, 2):
for dw in range(-1, 2):
_h = cluster.h + dh
_w = cluster.w + dw
new_gradient = self.get_gradient(_h, _w)
if new_gradient < cluster_gradient:
cluster.update(_h, _w, self.data[_h][_w][0], self.data[_h][_w][1], self.data[_h][_w][2])
cluster_gradient = new_gradient
def assignment(self):
for cluster in self.clusters:
for h in range(cluster.h - 2 * self.S, cluster.h + 2 * self.S):
if h < 0 or h >= self.image_height: continue
for w in range(cluster.w - 2 * self.S, cluster.w + 2 * self.S):
if w < 0 or w >= self.image_width: continue
L, A, B = self.data[h][w]
Dc = math.sqrt(
math.pow(L - cluster.l, 2) +
math.pow(A - cluster.a, 2) +
math.pow(B - cluster.b, 2))
Ds = math.sqrt(
math.pow(h - cluster.h, 2) +
math.pow(w - cluster.w, 2))
D = math.sqrt(math.pow(Dc / self.M, 2) + math.pow(Ds / self.S, 2))
if D < self.dis[h][w]:
if (h, w) not in self.label:
self.label[(h, w)] = cluster
cluster.pixels.append((h, w))
else:
self.label[(h, w)].pixels.remove((h, w))
self.label[(h, w)] = cluster
cluster.pixels.append((h, w))
self.dis[h][w] = D
def update_cluster(self):
for cluster in self.clusters:
sum_h = sum_w = number =0
for p in cluster.pixels:
sum_h += p[0]
sum_w += p[1]
number += 1
_h = int(sum_h / number)
_w = int(sum_w / number)
cluster.update(_h, _w, self.data[_h][_w][0], self.data[_h][_w][1], self.data[_h][_w][2])
def getGlcm(self,d_y,d_x):#灰度共生矩阵
#将一整张图片转化为灰度图
glcm_img = cv2.imread(f"C:/Users/Administrator/Desktop/SRP/Dataset_BUSI_with_GT/benign/benign ({self.n}).png", 0)
for cluster in self.clusters: #每个超像素块
max_gray_level = number=0
ret = [[0.0 for i in range(gray_level)] for j in range(gray_level)]
for p in cluster.pixels:
number += 1
if glcm_img[p[0]][p[1]] > max_gray_level:
max_gray_level = glcm_img[p[0]][p[1]]
max_gray_level = max_gray_level + 1 #得到此超像素块中最大灰度级
for pix in cluster.pixels:
if max_gray_level > gray_level: #若是最大灰度级大于设定灰度级，则将其调整为设定灰度级大小
glcm_img[pix[0]][pix[1]] = glcm_img[pix[0]][pix[1]] * gray_level / max_gray_level
for pixe in cluster.pixels: #再次逐像素遍历
h_1 = pixe[0] + d_y #设定边界
w_1 = pixe[1] + d_x
if h_1 > glcm_img.shape[0] - 1:
h_1 = glcm_img.shape[0] - 1
if w_1 > glcm_img.shape[1] - 1:
w_1 = glcm_img.shape[1] - 1
rows = glcm_img[pixe[0]][pixe[1]]
cols = glcm_img[h_1][w_1]
if rows >= 16:
rows = 15
if cols >= 16:
cols = 15
ret[rows][cols] += 1.0
for i in range(gray_level):
for j in range(gray_level):
ret[i][j] /= float(number) #得到灰度生成矩阵
asm, con, eng, idm = feature_computer(ret)
string=str([asm, con, eng, idm])
with open(f'C:/Users/Administrator/Desktop/data/{self.n}/benign{self.n}_d_y={d_y}_d_x={d_x}.txt', 'a') as f:
f.write('\n'+ string)
def average(self):#数据平均值
for cluster in self.clusters: #每一个超像素
l = a = b = number =0
for p in cluster.pixels: #超像素中的每一个像素
l += self.data[p[0]][p[1]][0]
a += self.data[p[0]][p[1]][1]
b += self.data[p[0]][p[1]][2]
number += 1 #像素总个数
strin=str([l/number,a/number,b/number]) #均值
with open(f'C:/Users/Administrator/Desktop/data/{self.n}/benign{self.n}_average.txt', 'a') as f: #保存数据
f.write('\n' + strin)
def fangcha(self):#数据方差
for cluster in self.clusters:
l = a = b = []
for p in cluster.pixels:
l.append(self.data[p[0]][p[1]][0])
a.append(self.data[p[0]][p[1]][1])
b.append(self.data[p[0]][p[1]][2])
stri=str([np.var(l),np.var(a),np.var(b)]) #方差
with open(f'C:/Users/Administrator/Desktop/data/{self.n}/benign{self.n}_variance.txt', 'a') as f:
f.write('\n' + stri)
def save_current_image(self, name):
image_arr = np.copy(self.data)
for cluster in self.clusters:
for p in cluster.pixels:
image_arr[p[0]][p[1]][0] = cluster.l
image_arr[p[0]][p[1]][1] = cluster.a
image_arr[p[0]][p[1]][2] = cluster.b
image_arr[cluster.h][cluster.w][0] = 0
image_arr[cluster.h][cluster.w][1] = 0
image_arr[cluster.h][cluster.w][2] = 0
self.save_lab_image(name, image_arr)
def iterate_10times(self):
self.init_clusters()
self.move_clusters()
for i in trange(10):
self.assignment()
self.update_cluster()
if i == 9:
self.getGlcm(0,1)#参数为d_y,d_x
self.getGlcm(-1,0)
self.getGlcm(1,0)
self.getGlcm(1,1)
self.getGlcm(1,-1)
self.getGlcm(-1,1)
self.getGlcm(-1,-1)
self.average()#平均值
self.fangcha()#方差
name = 'benign{n}_lenna_M{m}_K{k}_loop{loop}.png'.format(n=self.n,loop=i, m=self.M, k=self.K)
self.save_current_image(name)
if __name__ == '__main__':
for i in range(110,135):
p = SLICProcessor(f'C:/Users/Administrator/Desktop/SRP/Dataset_BUSI_with_GT/benign/benign ({i}).png', 200, 30,i)
p.iterate_10times()本文章为转载内容，我们尊重原作者对文章享有的著作权。如有内容错误或侵权问题，欢迎原作者联系我们进行内容更正或删除文章。
