-
Notifications
You must be signed in to change notification settings - Fork 0
/
LBP.py
65 lines (55 loc) · 1.91 KB
/
LBP.py
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
#coding:utf-8
# 入出力の為のライブラリ #
import cv2
import copy
import math
def main():
#--- 画像入力処理 ---#
print ("Please input picture")
input_picture = raw_input('>>> ')
input_img = cv2.imread(input_picture,0)
lbp_img = copy.deepcopy(input_img)
img_hor = len(input_img)
img_ver = len(input_img[0])
histgram = [0 for col in range(256)]
points=8
radius=1
#--- LBP処理部分 ---#
for i in range(0,img_hor):
for j in range(0,img_ver):
# reset array
mask=[0 for col in range(points)]
lbp_img[i][j]=2**8-1
# when ( left / right / top / bottom ) exist, calc LBP
if i>radius and i+radius<img_ver and j>radius and j+radius<img_hor :
#--- LBP算出部分 ---#
for part in range (0,points):
# measure distance between [j][j] and its neighborhoods
x_len = int(radius * math.cos(360/points * part)+0.5)
y_len = int(radius * math.sin(360/points * part)+0.5)
# get mask pattern -正負判定はsign()でもok-
if input_img[i+x_len][j+y_len] - input_img[i][j] > 0:
mask[part] =1
# loop time to find min value
print(mask)
for k in range (0,points):
lbp_val=0
for x in range (0+k,points):
lbp_val += mask[x]* (2**(x-k))
for y in range (0,k):
lbp_val += mask[y]* (2**(y+(points-k)))
lbp_img[i][j]=min(lbp_img[i][j],lbp_val)
print (lbp_img[i][j])
#--- ヒストグラム作成 ---#
for i in range(0,img_hor):
for j in range(0,img_ver):
histgram[lbp_img[i][j]] += 1
#--- 結果表示部分 ---#
cv2.imshow("Local Binary Pattern",lbp_img)
# when input key was 's', save result image or fin
if cv2.waitKey(0) == ord('s'):
cv2.imwrite("LBP.png", lbp_img)
cv2.destroyAllWindow()
# fin main function
if __name__ == '__main__':
main()