Skip to content

Commit 9614e2b

Browse files
LauLauThomLauLauThom
authored
Merge pull request #1 from multi-template-matching/use-Pandas
Use pandas
2 parents 2de08b7 + 9fc1c94 commit 9614e2b

File tree

6 files changed

+329
-303
lines changed

6 files changed

+329
-303
lines changed

‎MTM/NMS.py

Lines changed: 175 additions & 178 deletions
Original file line numberDiff line numberDiff line change
@@ -11,192 +11,189 @@
1111
@author: Laurent Thomas
1212
"""
1313
from __future__ import division, print_function # for compatibility with Py2
14+
import pandas as pd
1415

1516
def Point_in_Rectangle(Point, Rectangle):
16-
'''Return True if a point (x,y) is contained in a Rectangle(x, y, width, height)'''
17-
# unpack variables
18-
Px, Py = Point
19-
Rx, Ry, w, h = Rectangle
17+
'''Return True if a point (x,y) is contained in a Rectangle(x, y, width, height)'''
18+
# unpack variables
19+
Px, Py = Point
20+
Rx, Ry, w, h = Rectangle
2021

21-
return (Rx <= Px) and (Px <= Rx + w -1) and (Ry <= Py) and (Py <= Ry + h -1) # simply test if x_Point is in the range of x for the rectangle
22+
return (Rx <= Px) and (Px <= Rx + w -1) and (Ry <= Py) and (Py <= Ry + h -1) # simply test if x_Point is in the range of x for the rectangle
2223

2324

2425
def computeIoU(BBox1,BBox2):
25-
'''
26-
Compute the IoU (Intersection over Union) between 2 rectangular bounding boxes defined by the top left (Xtop,Ytop) and bottom right (Xbot, Ybot) pixel coordinates
27-
Code adapted from https://www.pyimagesearch.com/2016/11/07/intersection-over-union-iou-for-object-detection/
28-
'''
29-
#print('BBox1 : ', BBox1)
30-
#print('BBox2 : ', BBox2)
31-
32-
# Unpack input (python3 - tuple are no more supported as input in function definition - PEP3113 - Tuple can be used in as argument in a call but the function will not unpack it automatically)
33-
Xleft1, Ytop1, Width1, Height1 = BBox1
34-
Xleft2, Ytop2, Width2, Height2 = BBox2
35-
36-
# Compute bottom coordinates
37-
Xright1 = Xleft1 + Width1 -1 # we remove -1 from the width since we start with 1 pixel already (the top one)
38-
Ybot1 = Ytop1 + Height1 -1 # idem for the height
39-
40-
Xright2 = Xleft2 + Width2 -1
41-
Ybot2 = Ytop2 + Height2 -1
42-
43-
# determine the (x, y)-coordinates of the top left and bottom right points of the intersection rectangle
44-
Xleft = max(Xleft1, Xleft2)
45-
Ytop = max(Ytop1, Ytop2)
46-
Xright = min(Xright1, Xright2)
47-
Ybot = min(Ybot1, Ybot2)
48-
49-
# Compute boolean for inclusion
50-
BBox1_in_BBox2 = Point_in_Rectangle((Xleft1, Ytop1), BBox2) and Point_in_Rectangle((Xleft1, Ybot1), BBox2) and Point_in_Rectangle((Xright1, Ytop1), BBox2) and Point_in_Rectangle((Xright1, Ybot1), BBox2)
51-
BBox2_in_BBox1 = Point_in_Rectangle((Xleft2, Ytop2), BBox1) and Point_in_Rectangle((Xleft2, Ybot2), BBox1) and Point_in_Rectangle((Xright2, Ytop2), BBox1) and Point_in_Rectangle((Xright2, Ybot2), BBox1)
52-
53-
# Check that for the intersection box, Xtop,Ytop is indeed on the top left of Xbot,Ybot
54-
if BBox1_in_BBox2 or BBox2_in_BBox1:
55-
#print('One BBox is included within the other')
56-
IoU = 1
57-
58-
elif Xright<Xleft or Ybot<Ytop : # it means that there is no intersection (bbox is inverted)
59-
#print('No overlap')
60-
IoU = 0
61-
62-
else:
63-
# Compute area of the intersecting box
64-
Inter = (Xright - Xleft + 1) * (Ybot - Ytop + 1) # +1 since we are dealing with pixels. See a 1D example with 3 pixels for instance
65-
#print('Intersection area : ', Inter)
66-
67-
# Compute area of the union as Sum of the 2 BBox area - Intersection
68-
Union = Width1 * Height1 + Width2 * Height2 - Inter
69-
#print('Union : ', Union)
70-
71-
# Compute Intersection over union
72-
IoU = Inter/Union
73-
74-
#print('IoU : ',IoU)
75-
return IoU
76-
77-
78-
79-
def NMS(List_Hit, scoreThreshold=None, sortDescending=True, N_object=float("inf"), maxOverlap=0.5):
80-
'''
81-
Perform Non-Maxima supression : it compares the hits after maxima/minima detection, and removes the ones that are too close (too large overlap)
82-
This function works both with an optionnal threshold on the score, and number of detected bbox
83-
84-
if a scoreThreshold is specified, we first discard any hit below/above the threshold (depending on sortDescending)
85-
if sortDescending = True, the hit with score below the treshold are discarded (ie when high score means better prediction ex : Correlation)
86-
if sortDescending = False, the hit with score above the threshold are discared (ie when low score means better prediction ex : Distance measure)
87-
88-
Then the hit are ordered so that we have the best hits first.
89-
Then we iterate over the list of hits, taking one hit at a time and checking for overlap with the previous validated hit (the Final Hit list is directly iniitialised with the first best hit as there is no better hit with which to compare overlap)
90-
91-
This iteration is terminate once we have collected N best hit, or if there are no more hit left to test for overlap
26+
'''
27+
Compute the IoU (Intersection over Union) between 2 rectangular bounding boxes defined by the top left (Xtop,Ytop) and bottom right (Xbot, Ybot) pixel coordinates
28+
Code adapted from https://www.pyimagesearch.com/2016/11/07/intersection-over-union-iou-for-object-detection/
29+
'''
30+
#print('BBox1 : ', BBox1)
31+
#print('BBox2 : ', BBox2)
32+
33+
# Unpack input (python3 - tuple are no more supported as input in function definition - PEP3113 - Tuple can be used in as argument in a call but the function will not unpack it automatically)
34+
Xleft1, Ytop1, Width1, Height1 = BBox1
35+
Xleft2, Ytop2, Width2, Height2 = BBox2
36+
37+
# Compute bottom coordinates
38+
Xright1 = Xleft1 + Width1 -1 # we remove -1 from the width since we start with 1 pixel already (the top one)
39+
Ybot1 = Ytop1 + Height1 -1 # idem for the height
40+
41+
Xright2 = Xleft2 + Width2 -1
42+
Ybot2 = Ytop2 + Height2 -1
43+
44+
# determine the (x, y)-coordinates of the top left and bottom right points of the intersection rectangle
45+
Xleft = max(Xleft1, Xleft2)
46+
Ytop = max(Ytop1, Ytop2)
47+
Xright = min(Xright1, Xright2)
48+
Ybot = min(Ybot1, Ybot2)
49+
50+
# Compute boolean for inclusion
51+
BBox1_in_BBox2 = Point_in_Rectangle((Xleft1, Ytop1), BBox2) and Point_in_Rectangle((Xleft1, Ybot1), BBox2) and Point_in_Rectangle((Xright1, Ytop1), BBox2) and Point_in_Rectangle((Xright1, Ybot1), BBox2)
52+
BBox2_in_BBox1 = Point_in_Rectangle((Xleft2, Ytop2), BBox1) and Point_in_Rectangle((Xleft2, Ybot2), BBox1) and Point_in_Rectangle((Xright2, Ytop2), BBox1) and Point_in_Rectangle((Xright2, Ybot2), BBox1)
53+
54+
# Check that for the intersection box, Xtop,Ytop is indeed on the top left of Xbot,Ybot
55+
if BBox1_in_BBox2 or BBox2_in_BBox1:
56+
#print('One BBox is included within the other')
57+
IoU = 1
58+
59+
elif Xright<Xleft or Ybot<Ytop : # it means that there is no intersection (bbox is inverted)
60+
#print('No overlap')
61+
IoU = 0
62+
63+
else:
64+
# Compute area of the intersecting box
65+
Inter = (Xright - Xleft + 1) * (Ybot - Ytop + 1) # +1 since we are dealing with pixels. See a 1D example with 3 pixels for instance
66+
#print('Intersection area : ', Inter)
67+
68+
# Compute area of the union as Sum of the 2 BBox area - Intersection
69+
Union = Width1 * Height1 + Width2 * Height2 - Inter
70+
#print('Union : ', Union)
71+
72+
# Compute Intersection over union
73+
IoU = Inter/Union
74+
75+
#print('IoU : ',IoU)
76+
return IoU
77+
78+
79+
80+
def NMS(tableHit, scoreThreshold=None, sortAscending=False, N_object=float("inf"), maxOverlap=0.5):
81+
'''
82+
Perform Non-Maxima supression : it compares the hits after maxima/minima detection, and removes the ones that are too close (too large overlap)
83+
This function works both with an optionnal threshold on the score, and number of detected bbox
84+
85+
if a scoreThreshold is specified, we first discard any hit below/above the threshold (depending on sortDescending)
86+
if sortDescending = True, the hit with score below the treshold are discarded (ie when high score means better prediction ex : Correlation)
87+
if sortDescending = False, the hit with score above the threshold are discared (ie when low score means better prediction ex : Distance measure)
88+
89+
Then the hit are ordered so that we have the best hits first.
90+
Then we iterate over the list of hits, taking one hit at a time and checking for overlap with the previous validated hit (the Final Hit list is directly iniitialised with the first best hit as there is no better hit with which to compare overlap)
91+
92+
This iteration is terminate once we have collected N best hit, or if there are no more hit left to test for overlap
9293
9394
INPUT
94-
- ListHit : a list of dictionnary, with each dictionnary being a hit following the formating {'TemplateIdx'= (int),'BBox'=(x,y,width,height),'Score'=(float)}
95-
the TemplateIdx is the row index in the panda/Knime table
96-
97-
- scoreThreshold : Float (or None), used to remove hit with too low prediction score.
98-
If sortDescending=True (ie we use a correlation measure so we want to keep large scores) the scores above that threshold are kept
99-
While if we use sortDescending=False (we use a difference measure ie we want to keep low score), the scores below that threshold are kept
100-
101-
- N_object : number of best hit to return (by increasing score). Min=1, eventhough it does not really make sense to do NMS with only 1 hit
102-
- maxOverlap : float between 0 and 1, the maximal overlap authorised between 2 bounding boxes, above this value, the bounding box of lower score is deleted
103-
- sortDescending : use True when high score means better prediction, False otherwise (ex : if score is a difference measure, then the best prediction are low difference and we sort by ascending order)
104-
105-
OUTPUT
106-
List_nHit : List of the best detection after NMS, it contains max N detection (but potentially less)
107-
'''
108-
109-
# Apply threshold on prediction score
110-
if scoreThreshold==None :
111-
List_ThreshHit = List_Hit[:] # copy to avoid modifying the input list in place
112-
113-
elif sortDescending : # We keep hit above the threshold
114-
List_ThreshHit = [dico for dico in List_Hit if dico['Score']>=scoreThreshold]
115-
116-
elif not sortDescending : # We keep hit below the threshold
117-
List_ThreshHit = [dico for dico in List_Hit if dico['Score']<=scoreThreshold]
118-
119-
120-
# Sort score to have best predictions first (important as we loop testing the best boxes against the other boxes)
121-
if sortDescending:
122-
List_ThreshHit.sort(key=lambda dico: dico['Score'], reverse=True) # Hit = [list of (x,y),score] - sort according to descending (best = high correlation)
123-
else:
124-
List_ThreshHit.sort(key=lambda dico: dico['Score']) # sort according to ascending score (best = small difference)
125-
126-
127-
# Split the inital pool into Final Hit that are kept and restHit that can be tested
128-
# Initialisation : 1st keep is kept for sure, restHit is the rest of the list
129-
#print("\nInitialise final hit list with first best hit")
130-
FinalHit = [List_ThreshHit[0]]
131-
restHit = List_ThreshHit[1:]
132-
133-
134-
# Loop to compute overlap
135-
while len(FinalHit)<N_object and restHit : # second condition is restHit is not empty
136-
137-
# Report state of the loop
138-
#print("\n\n\nNext while iteration")
139-
140-
#print("-> Final hit list")
141-
#for hit in FinalHit: print(hit)
142-
143-
#print("\n-> Remaining hit list")
144-
#for hit in restHit: print(hit)
145-
146-
# pick the next best peak in the rest of peak
147-
test_hit = restHit[0]
148-
test_bbox = test_hit['BBox']
149-
#print("\nTest BBox:{} for overlap against higher score bboxes".format(test_bbox))
150-
151-
# Loop over hit in FinalHit to compute successively overlap with test_peak
152-
for hit in FinalHit:
153-
154-
# Recover Bbox from hit
155-
bbox2 = hit['BBox']
156-
157-
# Compute the Intersection over Union between test_peak and current peak
158-
IoU = computeIoU(test_bbox, bbox2)
159-
160-
# Initialise the boolean value to true before test of overlap
161-
ToAppend = True
162-
163-
if IoU>maxOverlap:
164-
ToAppend = False
165-
#print("IoU above threshold\n")
166-
break # no need to test overlap with the other peaks
167-
168-
else:
169-
#print("IoU below threshold\n")
170-
# no overlap for this particular (test_peak,peak) pair, keep looping to test the other (test_peak,peak)
171-
continue
172-
173-
174-
# After testing against all peaks (for loop is over), append or not the peak to final
175-
if ToAppend:
176-
# Move the test_hit from restHit to FinalHit
177-
#print("Append {} to list of final hits, remove it from Remaining hit list".format(test_hit))
178-
FinalHit.append(test_hit)
179-
restHit.remove(test_hit)
180-
181-
else:
182-
# only remove the test_peak from restHit
183-
#print("Remove {} from Remaining hit list".format(test_hit))
184-
restHit.remove(test_hit)
185-
186-
187-
# Once function execution is done, return list of hit without overlap
188-
#print("\nCollected N expected hit, or no hit left to test")
189-
#print("NMS over\n")
190-
return FinalHit
191-
192-
95+
- tableHit : (Panda DataFrame) Each row is a hit, with columns "TemplateName"(String),"BBox"(x,y,width,height),"Score"(float)
96+
97+
- scoreThreshold : Float (or None), used to remove hit with too low prediction score.
98+
If sortDescending=True (ie we use a correlation measure so we want to keep large scores) the scores above that threshold are kept
99+
While if we use sortDescending=False (we use a difference measure ie we want to keep low score), the scores below that threshold are kept
100+
101+
- N_object : number of best hit to return (by increasing score). Min=1, eventhough it does not really make sense to do NMS with only 1 hit
102+
- maxOverlap : float between 0 and 1, the maximal overlap authorised between 2 bounding boxes, above this value, the bounding box of lower score is deleted
103+
- sortAscending : use True when low score means better prediction (Difference-based score), True otherwise (Correlation score)
104+
105+
OUTPUT
106+
Panda DataFrame with best detection after NMS, it contains max N detection (but potentially less)
107+
'''
108+
109+
# Apply threshold on prediction score
110+
if scoreThreshold==None :
111+
threshTable = tableHit.copy() # copy to avoid modifying the input list in place
112+
113+
elif not sortAscending : # We keep rows above the threshold
114+
threshTable = tableHit[ tableHit['Score']>=scoreThreshold ]
115+
116+
elif sortAscending : # We keep hit below the threshold
117+
threshTable = tableHit[ tableHit['Score']<=scoreThreshold ]
118+
119+
# Sort score to have best predictions first (ie lower score if difference-based, higher score if correlation-based)
120+
# important as we loop testing the best boxes against the other boxes)
121+
threshTable.sort_values("Score", ascending=sortAscending, inplace=True) # Warning here is fine
122+
123+
124+
# Split the inital pool into Final Hit that are kept and restTable that can be tested
125+
# Initialisation : 1st keep is kept for sure, restTable is the rest of the list
126+
#print("\nInitialise final hit list with first best hit")
127+
outTable = threshTable.iloc[[0]].to_dict('records') # double square bracket to recover a DataFrame
128+
restTable = threshTable.iloc[1:].to_dict('records')
129+
130+
131+
# Loop to compute overlap
132+
while len(outTable)<N_object and restTable: # second condition is restTable is not empty
133+
134+
# Report state of the loop
135+
#print("\n\n\nNext while iteration")
136+
137+
#print("-> Final hit list")
138+
#for hit in outTable: print(hit)
139+
140+
#print("\n-> Remaining hit list")
141+
#for hit in restTable: print(hit)
142+
143+
# pick the next best peak in the rest of peak
144+
testHit_dico = restTable[0] # dico
145+
test_bbox = testHit_dico['BBox']
146+
#print("\nTest BBox:{} for overlap against higher score bboxes".format(test_bbox))
147+
148+
# Loop over hit in outTable to compute successively overlap with testHit
149+
for hit_dico in outTable:
150+
151+
# Recover Bbox from hit
152+
bbox2 = hit_dico['BBox']
153+
154+
# Compute the Intersection over Union between test_peak and current peak
155+
IoU = computeIoU(test_bbox, bbox2)
156+
157+
# Initialise the boolean value to true before test of overlap
158+
ToAppend = True
159+
160+
if IoU>maxOverlap:
161+
ToAppend = False
162+
#print("IoU above threshold\n")
163+
break # no need to test overlap with the other peaks
164+
165+
else:
166+
#print("IoU below threshold\n")
167+
# no overlap for this particular (test_peak,peak) pair, keep looping to test the other (test_peak,peak)
168+
continue
169+
170+
171+
# After testing against all peaks (for loop is over), append or not the peak to final
172+
if ToAppend:
173+
# Move the test_hit from restTable to outTable
174+
#print("Append {} to list of final hits, remove it from Remaining hit list".format(test_hit))
175+
outTable.append(testHit_dico)
176+
restTable.remove(testHit_dico)
177+
178+
else:
179+
# only remove the test_peak from restTable
180+
#print("Remove {} from Remaining hit list".format(test_hit))
181+
restTable.remove(testHit_dico)
182+
183+
184+
# Once function execution is done, return list of hit without overlap
185+
#print("\nCollected N expected hit, or no hit left to test")
186+
#print("NMS over\n")
187+
return pd.DataFrame(outTable)
188+
189+
193190
if __name__ == "__main__":
194-
Hit1 = {'TemplateIdx':1,'BBox':(780, 350, 700, 480), 'Score':0.8}
195-
Hit2 = {'TemplateIdx':1,'BBox':(806, 416, 716, 442), 'Score':0.6}
196-
Hit3 = {'TemplateIdx':1,'BBox':(1074, 530, 680, 390), 'Score':0.4}
191+
ListHit =[
192+
{'TemplateName':1,'BBox':(780, 350, 700, 480), 'Score':0.8},
193+
{'TemplateName':1,'BBox':(806, 416, 716, 442), 'Score':0.6},
194+
{'TemplateName':1,'BBox':(1074, 530, 680, 390), 'Score':0.4}
195+
]
197196

198-
ListHit = [Hit1, Hit2, Hit3]
197+
FinalHits = NMS( pd.DataFrame(ListHit), scoreThreshold=0.7, sortAscending=False, maxOverlap=0.5 )
199198

200-
ListFinalHit = NMS(ListHit)
201-
202-
print(ListFinalHit)
199+
print(FinalHits)

0 commit comments

Comments
 (0)