class TextProposalConnector:
"""
Connect text proposals into text lines
"""
def __init__(self):
self.graph_builder = TextProposalGraphBuilder()
def group_text_proposals(self, text_proposals, scores, im_size):
graph = self.graph_builder.build_graph(text_proposals, scores, im_size)
return graph.sub_graphs_connected()
def fit_y(self, X, Y, x1, x2):
len(X) != 0
# if X only include one point, the function will get line y=Y[0]
if np.sum(X == X[0]) == len(X):
return Y[0], Y[0]
p = np.poly1d(np.polyfit(X, Y, 1))
return p(x1), p(x2)
def get_text_lines(self, text_proposals, scores, im_size):
# tp=text proposal
tp_groups = self.group_text_proposals(text_proposals, scores, im_size)
text_lines = np.zeros((len(tp_groups), 8), np.float32)
for index, tp_indices in enumerate(tp_groups):
text_line_boxes = text_proposals[list(tp_indices)]
num = np.size(text_line_boxes)
X = (text_line_boxes[:, 0] + text_line_boxes[:, 2]) / 2
Y = (text_line_boxes[:, 1] + text_line_boxes[:, 3]) / 2
z1 = np.polyfit(X, Y, 1)
p1 = np.poly1d(z1)
x0 = np.min(text_line_boxes[:, 0])
x1 = np.max(text_line_boxes[:, 2])
offset = (text_line_boxes[0, 2] - text_line_boxes[0, 0]) * 0.5
lt_y, rt_y = self.fit_y(text_line_boxes[:, 0], text_line_boxes[:,
1],
x0 + offset, x1 - offset)
lb_y, rb_y = self.fit_y(text_line_boxes[:, 0], text_line_boxes[:,
3],
x0 + offset, x1 - offset)
# the score of a text line is the average score of the scores
# of all text proposals contained in the text line
score = scores[list(tp_indices)].sum() / float(len(tp_indices))
text_lines[index, 0] = x0
text_lines[index, 1] = min(lt_y, rt_y)
text_lines[index, 2] = x1
text_lines[index, 3] = max(lb_y, rb_y)
text_lines[index, 4] = score
text_lines[index, 5] = z1[0]
text_lines[index, 6] = z1[1]
height = np.mean((text_line_boxes[:, 3] - text_line_boxes[:, 1]))
text_lines[index, 7] = height + 2.5
#text_lines=clip_boxes(text_lines, im_size)
return text_lines
class TextProposalConnector:
def __init__(self):
self.graph_builder=TextProposalGraphBuilder()
def group_text_proposals(self, text_proposals, scores, im_size):
graph=self.graph_builder.build_graph(text_proposals, scores, im_size)
return graph.sub_graphs_connected()
def fit_y(self, X, Y, x1, x2):
len(X)!=0
# if X only include one point, the function will get line y=Y[0]
if np.sum(X==X[0])==len(X):
return Y[0], Y[0]
p=np.poly1d(np.polyfit(X, Y, 1))
return p(x1), p(x2)
def get_text_lines(self, text_proposals, scores, im_size):
# tp=text proposal
tp_groups=self.group_text_proposals(text_proposals, scores, im_size)
text_lines=np.zeros((len(tp_groups), 5), np.float32)
for index, tp_indices in enumerate(tp_groups):
text_line_boxes=text_proposals[list(tp_indices)]
x0=np.min(text_line_boxes[:, 0])
x1=np.max(text_line_boxes[:, 2])
offset=(text_line_boxes[0, 2]-text_line_boxes[0, 0])*0.5
lt_y, rt_y=self.fit_y(text_line_boxes[:, 0], text_line_boxes[:, 1], x0+offset, x1-offset)
lb_y, rb_y=self.fit_y(text_line_boxes[:, 0], text_line_boxes[:, 3], x0+offset, x1-offset)
# the score of a text line is the average score of the scores
# of all text proposals contained in the text line
score=scores[list(tp_indices)].sum()/float(len(tp_indices))
text_lines[index, 0]=x0
text_lines[index, 1]=min(lt_y, rt_y)
text_lines[index, 2]=x1
text_lines[index, 3]=max(lb_y, rb_y)
text_lines[index, 4]=score
text_lines=clip_boxes(text_lines, im_size)
text_recs = np.zeros((len(text_lines), 9), np.float)
index = 0
for line in text_lines:
xmin,ymin,xmax,ymax=line[0],line[1],line[2],line[3]
text_recs[index, 0] = xmin
text_recs[index, 1] = ymin
text_recs[index, 2] = xmax
text_recs[index, 3] = ymin
text_recs[index, 4] = xmin
text_recs[index, 5] = ymax
text_recs[index, 6] = xmax
text_recs[index, 7] = ymax
text_recs[index, 8] = line[4]
index = index + 1
return text_recs
class TextProposalConnector:
"""
Connect text proposals into text lines
"""
def __init__(self):
self.graph_builder=TextProposalGraphBuilder()
def group_text_proposals(self, text_proposals, scores, im_size):
graph=self.graph_builder.build_graph(text_proposals, scores, im_size)
return graph.sub_graphs_connected()
def fit_y(self, X, Y, x1, x2):
len(X)!=0
# if X only include one point, the function will get line y=Y[0]
if np.sum(X==X[0])==len(X):
return Y[0], Y[0]
p=np.poly1d(np.polyfit(X, Y, 1))
return p(x1), p(x2)
def get_text_lines(self, text_proposals, scores, im_size):
# tp=text proposal
tp_groups=self.group_text_proposals(text_proposals, scores, im_size)
text_lines=np.zeros((len(tp_groups), 8), np.float32)
for index, tp_indices in enumerate(tp_groups):
text_line_boxes=text_proposals[list(tp_indices)]
num = np.size(text_line_boxes)
X = (text_line_boxes[:,0] + text_line_boxes[:,2]) / 2
Y = (text_line_boxes[:,1] + text_line_boxes[:,3]) / 2
z1 = np.polyfit(X,Y,1)
p1 = np.poly1d(z1)
x0=np.min(text_line_boxes[:, 0])
x1=np.max(text_line_boxes[:, 2])
offset=(text_line_boxes[0, 2]-text_line_boxes[0, 0])*0.5
lt_y, rt_y=self.fit_y(text_line_boxes[:, 0], text_line_boxes[:, 1], x0+offset, x1-offset)
lb_y, rb_y=self.fit_y(text_line_boxes[:, 0], text_line_boxes[:, 3], x0+offset, x1-offset)
# the score of a text line is the average score of the scores
# of all text proposals contained in the text line
score=scores[list(tp_indices)].sum()/float(len(tp_indices))
text_lines[index, 0]=x0
text_lines[index, 1]=min(lt_y, rt_y)
text_lines[index, 2]=x1
text_lines[index, 3]=max(lb_y, rb_y)
text_lines[index, 4]=score
text_lines[index, 5]=z1[0]
text_lines[index, 6]=z1[1]
height = np.mean( (text_line_boxes[:,3]-text_line_boxes[:,1]) )
text_lines[index, 7]= height + 2.5
#text_lines=clip_boxes(text_lines, im_size)
return text_lines
class TextProposalConnector:
"""
Connect text proposals into text lines
"""
def __init__(self):
self.graph_builder = TextProposalGraphBuilder()
def group_text_proposals(self, text_proposals, scores, im_size):
graph = self.graph_builder.build_graph(text_proposals, scores, im_size)
return graph.sub_graphs_connected()
def fit_y(self, X, Y, x1, x2):
len(X) != 0
# if X only include one point, the function will get line y=Y[0]
if np.sum(X == X[0]) == len(X):
return Y[0], Y[0]
p = np.poly1d(np.polyfit(X, Y, 1))
return p(x1), p(x2)
#leastaq
def get_text_lines_leastaq(self, text_proposals, scores, im_size):
tp_groups = self.group_text_proposals(text_proposals, scores, im_size)
text_lines = np.zeros((len(tp_groups), 9), np.float32)
for index, tp_indices in enumerate(tp_groups):
text_line_boxes = text_proposals[list(tp_indices)]
x0 = np.min(text_line_boxes[:, 0])
x1 = np.max(text_line_boxes[:, 2])
p0 = [x0, x1]
Para1 = leastsq(error,
p0,
args=(np.array(text_line_boxes[:, 0]),
np.array(text_line_boxes[:, 1])))
Para2 = leastsq(error,
p0,
args=(np.array(text_line_boxes[:, 2]),
np.array(text_line_boxes[:, 3])))
k1, b1 = Para1[0]
k2, b2 = Para2[0]
y10 = k1 * x0 + b1
y11 = k1 * x1 + b1
y20 = k2 * x0 + b2
y21 = k2 * x1 + b2
score = scores[list(tp_indices)].sum() / float(len(tp_indices))
text_lines[index, 0] = x0
text_lines[index, 1] = y10
text_lines[index, 2] = x1
text_lines[index, 3] = y11
text_lines[index, 4] = x0
text_lines[index, 5] = y20
text_lines[index, 6] = x1
text_lines[index, 7] = y21
text_lines[index, 8] = score
return text_lines
def get_text_lines(self, text_proposals, scores, im_size):
# tp=text proposal
tp_groups = self.group_text_proposals(text_proposals, scores, im_size)
text_lines = np.zeros((len(tp_groups), 5), np.float32)
for index, tp_indices in enumerate(tp_groups):
text_line_boxes = text_proposals[list(tp_indices)]
x0 = np.min(text_line_boxes[:, 0])
x1 = np.max(text_line_boxes[:, 2])
offset = (text_line_boxes[0, 2] - text_line_boxes[0, 0]) * 0.5
lt_y, rt_y = self.fit_y(text_line_boxes[:, 0], text_line_boxes[:,
1],
x0 + offset, x1 - offset)
lb_y, rb_y = self.fit_y(text_line_boxes[:, 0], text_line_boxes[:,
3],
x0 + offset, x1 - offset)
# the score of a text line is the average score of the scores
# of all text proposals contained in the text line
score = scores[list(tp_indices)].sum() / float(len(tp_indices))
text_lines[index, 0] = x0
text_lines[index, 1] = min(lt_y, rt_y)
text_lines[index, 2] = x1
text_lines[index, 3] = max(lb_y, rb_y)
text_lines[index, 4] = score
text_lines = clip_boxes(text_lines, im_size)
return text_lines
class TextProposalConnectorOriented:
"""
Connect text proposals into text lines
"""
def __init__(self):
self.graph_builder = TextProposalGraphBuilder()
def group_text_proposals(self, text_proposals, scores, im_size):
graph = self.graph_builder.build_graph(text_proposals, scores, im_size)
return graph.sub_graphs_connected()
def fit_y(self, X, Y, x1, x2):
len(X) != 0
# if X only include one point, the function will get line y=Y[0]
if np.sum(X == X[0]) == len(X):
return Y[0], Y[0]
p = np.poly1d(np.polyfit(X, Y, 1))
return p(x1), p(x2)
def get_text_lines(self, text_proposals, scores, im_size):
"""
text_proposals:boxes
"""
# tp=text proposal
tp_groups = self.group_text_proposals(text_proposals, scores,
im_size) #首先还是建图，获取到文本行由哪几个小框构成
text_lines = np.zeros((len(tp_groups), 8), np.float32)
for index, tp_indices in enumerate(tp_groups):
text_line_boxes = text_proposals[list(tp_indices)] #每个文本行的全部小框
X = (text_line_boxes[:, 0] +
text_line_boxes[:, 2]) / 2 # 求每一个小框的中心x，y坐标
Y = (text_line_boxes[:, 1] + text_line_boxes[:, 3]) / 2
z1 = np.polyfit(X, Y, 1) #多项式拟合，根据之前求的中心店拟合一条直线（最小二乘）
x0 = np.min(text_line_boxes[:, 0]) #文本行x坐标最小值
x1 = np.max(text_line_boxes[:, 2]) #文本行x坐标最大值
offset = (text_line_boxes[0, 2] -
text_line_boxes[0, 0]) * 0.5 #小框宽度的一半
# 以全部小框的左上角这个点去拟合一条直线，然后计算一下文本行x坐标的极左极右对应的y坐标
lt_y, rt_y = self.fit_y(text_line_boxes[:, 0], text_line_boxes[:,
1],
x0 + offset, x1 - offset)
# 以全部小框的左下角这个点去拟合一条直线，然后计算一下文本行x坐标的极左极右对应的y坐标
lb_y, rb_y = self.fit_y(text_line_boxes[:, 0], text_line_boxes[:,
3],
x0 + offset, x1 - offset)
score = scores[list(tp_indices)].sum() / float(
len(tp_indices)) #求全部小框得分的均值作为文本行的均值
text_lines[index, 0] = x0
text_lines[index, 1] = min(lt_y, rt_y) #文本行上端 线段 的y坐标的小值
text_lines[index, 2] = x1
text_lines[index, 3] = max(lb_y, rb_y) #文本行下端 线段 的y坐标的大值
text_lines[index, 4] = score #文本行得分
text_lines[index, 5] = z1[0] #根据中心点拟合的直线的k，b
text_lines[index, 6] = z1[1]
height = np.mean(
(text_line_boxes[:, 3] - text_line_boxes[:, 1])) #小框平均高度
text_lines[index, 7] = height + 2.5
text_recs = np.zeros((len(text_lines), 9), np.float)
index = 0
for line in text_lines:
b1 = line[6] - line[7] / 2 # 根据高度和文本行中心线，求取文本行上下两条线的b值
b2 = line[6] + line[7] / 2
x1 = line[0]
y1 = line[5] * line[0] + b1 # 左上
x2 = line[2]
y2 = line[5] * line[2] + b1 # 右上
x3 = line[0]
y3 = line[5] * line[0] + b2 # 左下
x4 = line[2]
y4 = line[5] * line[2] + b2 # 右下
disX = x2 - x1
disY = y2 - y1
width = np.sqrt(disX * disX + disY * disY) # 文本行宽度
fTmp0 = y3 - y1 # 文本行高度
fTmp1 = fTmp0 * disY / width
x = np.fabs(fTmp1 * disX / width) # 做补偿
y = np.fabs(fTmp1 * disY / width)
if line[5] < 0:
x1 -= x
y1 += y
x4 += x
y4 -= y
else:
x2 += x
y2 += y
x3 -= x
y3 -= y
text_recs[index, 0] = x1
text_recs[index, 1] = y1
text_recs[index, 2] = x2
text_recs[index, 3] = y2
text_recs[index, 4] = x3
text_recs[index, 5] = y3
text_recs[index, 6] = x4
text_recs[index, 7] = y4
text_recs[index, 8] = line[4]
index = index + 1
return text_recs
class TextProposalConnector:
"""
Connect text proposals into text lines
"""
def __init__(self):
self.graph_builder = TextProposalGraphBuilder()
def group_text_proposals(self, char_boxes, char_labels, im_size):
graph = self.graph_builder.build_graph(char_boxes, char_labels,
im_size)
return graph.sub_graphs_connected()
def fit_y(self, X, Y, x1, x2):
len(X) != 0
# if X only include one point, the function will get line y=Y[0]
if np.sum(X == X[0]) == len(X):
return Y[0], Y[0]
p = np.poly1d(np.polyfit(X, Y, 1))
return p(x1), p(x2)
def get_text_lines(self, char_boxes, char_labels, im_size):
# tp=text proposal
# the subgraph of graph, type is 'list of list '
tp_groups = self.group_text_proposals(char_boxes, char_labels, im_size)
text_lines = np.zeros((len(tp_groups), 4), np.float32)
text_lines_label = [''] * len(text_lines)
for index, tp_indices in enumerate(tp_groups):
text_line_boxes = char_boxes[list(tp_indices)]
x0 = np.min(text_line_boxes[:, 0])
x1 = np.max(text_line_boxes[:, 2])
offset = (text_line_boxes[0, 2] - text_line_boxes[0, 0]) * 0.5
#rectify the text proposals to a line which uses ployfit's method, then enlarge
lt_y, rt_y = self.fit_y(text_line_boxes[:, 0], text_line_boxes[:,
1],
x0 + offset, x1 - offset)
lb_y, rb_y = self.fit_y(text_line_boxes[:, 0], text_line_boxes[:,
3],
x0 + offset, x1 - offset)
# the score of a text line is the average score of the scores
# of all text proposals contained in the text line
y0 = min(lt_y, rt_y)
y1 = max(lb_y, rb_y)
if np.arctan(
(text_line_boxes[-1, 1] - y0) /
(text_line_boxes[-1, 2] - x0)) > pi * cfg.MAX_ANGLE / 180:
print 'current angle is', np.arctan(
(text_line_boxes[-1, 3] - y0) /
(text_line_boxes[-1, 2] - x0)) / pi * 180
x0, y0, x1, y1 = 0, 0, 0, 0
text_lines[index, 0] = x0
text_lines[index, 1] = y0
text_lines[index, 2] = x1
text_lines[index, 3] = y1
for tp_indice in tp_indices:
text_lines_label[index] += char_labels[tp_indice]
text_lines = clip_boxes(text_lines, im_size)
return text_lines, text_lines_label
class TextProposalConnector:
"""
Connect text proposals into text lines
"""
def __init__(self):
self.graph_builder = TextProposalGraphBuilder()
def group_text_proposals(self, text_proposals, scores, im_size):
graph = self.graph_builder.build_graph(text_proposals, scores, im_size)
return graph.sub_graphs_connected()
def fit_y(self, X, Y, x1, x2):
len(X) != 0
# if X only include one point, the function will get line y=Y[0]
if np.sum(X == X[0]) == len(X):
return Y[0], Y[0]
p = np.poly1d(np.polyfit(X, Y, 1))
return p(x1), p(x2)
def get_text_lines(self, text_proposals, scores, im_size, sides):
# tp=text proposal
tp_groups = self.group_text_proposals(text_proposals, scores, im_size)
text_lines = np.zeros((len(tp_groups), 5), np.float32)
for index, tp_indices in enumerate(tp_groups):
text_line_boxes = text_proposals[list(tp_indices)]
line_side = sides[list(tp_indices)]
x0 = np.min(text_line_boxes[:, 0])
x1 = np.max(text_line_boxes[:, 2])
for i in range(len(text_line_boxes)):
if text_line_boxes[i][0] == x0:
x0_index = i
if text_line_boxes[i][2] == x1:
x1_index = i
x0 = line_side[x0_index] * cfg.TEXT_PROPOSALS_WIDTH + (
text_line_boxes[x0_index][0] +
text_line_boxes[x0_index][2]) / 2.0
x1 = line_side[x1_index] * cfg.TEXT_PROPOSALS_WIDTH + (
text_line_boxes[x1_index][0] +
text_line_boxes[x1_index][2]) / 2.0
x0 = int(x0)
x1 = int(x1)
offset = (text_line_boxes[0, 2] - text_line_boxes[0, 0]) * 0.5
lt_y, rt_y = self.fit_y(text_line_boxes[:, 0], text_line_boxes[:,
1],
x0 + offset, x1 - offset)
lb_y, rb_y = self.fit_y(text_line_boxes[:, 0], text_line_boxes[:,
3],
x0 + offset, x1 - offset)
# the score of a text line is the average score of the scores
# of all text proposals contained in the text line
score = scores[list(tp_indices)].sum() / float(len(tp_indices))
text_lines[index, 0] = x0
text_lines[index, 1] = min(lt_y, rt_y)
text_lines[index, 2] = x1
text_lines[index, 3] = max(lb_y, rb_y)
text_lines[index, 4] = score
text_lines = clip_boxes(text_lines, im_size)
return text_lines
class TextProposalConnector:
"""
Connect text proposals into text lines
"""
def __init__(self):
self.graph_builder = TextProposalGraphBuilder()
def group_text_proposals(self, text_proposals, scores, im_size):
graph = self.graph_builder.build_graph(text_proposals, scores, im_size)
return graph.sub_graphs_connected()
def fit_y(self, X, Y, x1, x2):
len(X) != 0
# if X only include one point, the function will get line y=Y[0]
if np.sum(X == X[0]) == len(X):
return Y[0], Y[0]
p = np.poly1d(np.polyfit(X, Y, 1))
return p(x1), p(x2)
def get_text_lines(self, text_proposals, scores, xsides, im_size):
tp_groups = self.group_text_proposals(text_proposals, scores, im_size)
text_lines = np.zeros((len(tp_groups), 5), np.float32)
for index, tp_indices in enumerate(tp_groups):
text_line_boxes = text_proposals[list(tp_indices)]
left_anchor = text_proposals[tp_indices[0]]
right_anchor = text_proposals[tp_indices[-1]]
w_a_left = left_anchor[2] - left_anchor[0] + 1.
w_a_right = right_anchor[2] - right_anchor[0] + 1.
ctr_a_left = left_anchor[0] + w_a_left * 0.5
ctr_a_right = right_anchor[0] + w_a_right * 0.5
left_xside_pre = xsides[tp_indices[0]][0]
right_xside_pre = xsides[tp_indices[-1]][0]
x0_refined = left_xside_pre * w_a_left + ctr_a_left
x1_refined = right_xside_pre * w_a_right + ctr_a_right
x0 = np.min(text_line_boxes[:, 0])
x1 = np.max(text_line_boxes[:, 2])
offset = (text_line_boxes[0, 2] - text_line_boxes[0, 0]) * 0.5
lt_y, rt_y = self.fit_y(text_line_boxes[:, 0], text_line_boxes[:,
1],
x0_refined + offset, x1_refined - offset)
lb_y, rb_y = self.fit_y(text_line_boxes[:, 0], text_line_boxes[:,
3],
x0_refined + offset, x1_refined - offset)
# the score of a text line is the average score of the scores
# of all text proposals contained in the text line
score = scores[list(tp_indices)].sum() / float(len(tp_indices))
text_lines[index, 0] = x0_refined
text_lines[index, 1] = min(lt_y, rt_y)
text_lines[index, 2] = x1_refined
text_lines[index, 3] = max(lb_y, rb_y)
text_lines[index, 4] = score
text_lines = clip_boxes(text_lines, im_size)
return text_lines