mirror of https://github.com/open-mmlab/mmocr.git
113 lines
3.7 KiB
Python
113 lines
3.7 KiB
Python
# Copyright (c) OpenMMLab. All rights reserved.
|
|
import numpy as np
|
|
import pytest
|
|
import torch
|
|
|
|
from mmocr.models.textdet.postprocess import (DBPostprocessor,
|
|
FCEPostprocessor,
|
|
TextSnakePostprocessor)
|
|
from mmocr.models.textdet.postprocess.utils import comps2boundaries, poly_nms
|
|
|
|
|
|
def test_db_boxes_from_bitmaps():
|
|
"""Test the boxes_from_bitmaps function in db_decoder."""
|
|
pred = np.array([[[0.8, 0.8, 0.8, 0.8, 0], [0.8, 0.8, 0.8, 0.8, 0],
|
|
[0.8, 0.8, 0.8, 0.8, 0], [0.8, 0.8, 0.8, 0.8, 0],
|
|
[0.8, 0.8, 0.8, 0.8, 0]]])
|
|
preds = torch.FloatTensor(pred).requires_grad_(True)
|
|
db_decode = DBPostprocessor(text_repr_type='quad', min_text_width=0)
|
|
boxes = db_decode(preds)
|
|
assert len(boxes) == 1
|
|
|
|
|
|
def test_fcenet_decode():
|
|
|
|
k = 1
|
|
preds = []
|
|
preds.append(torch.ones(1, 4, 10, 10))
|
|
preds.append(torch.ones(1, 4 * k + 2, 10, 10))
|
|
fcenet_decode = FCEPostprocessor(
|
|
fourier_degree=k, num_reconstr_points=50, nms_thr=0.01)
|
|
boundaries = fcenet_decode(preds=preds, scale=1)
|
|
|
|
assert isinstance(boundaries, list)
|
|
|
|
|
|
def test_poly_nms():
|
|
threshold = 0
|
|
polygons = []
|
|
polygons.append([10, 10, 10, 30, 30, 30, 30, 10, 0.95])
|
|
polygons.append([15, 15, 15, 25, 25, 25, 25, 15, 0.9])
|
|
polygons.append([40, 40, 40, 50, 50, 50, 50, 40, 0.85])
|
|
polygons.append([5, 5, 5, 15, 15, 15, 15, 5, 0.7])
|
|
|
|
keep_poly = poly_nms(polygons, threshold)
|
|
assert isinstance(keep_poly, list)
|
|
|
|
|
|
def test_comps2boundaries():
|
|
|
|
# test comps2boundaries
|
|
x1 = np.arange(2, 18, 2)
|
|
x2 = x1 + 2
|
|
y1 = np.ones(8) * 2
|
|
y2 = y1 + 2
|
|
comp_scores = np.ones(8, dtype=np.float32) * 0.9
|
|
text_comps = np.stack([x1, y1, x2, y1, x2, y2, x1, y2,
|
|
comp_scores]).transpose()
|
|
comp_labels = np.array([1, 1, 1, 1, 1, 3, 5, 5])
|
|
shuffle = [3, 2, 5, 7, 6, 0, 4, 1]
|
|
boundaries = comps2boundaries(text_comps[shuffle], comp_labels[shuffle])
|
|
assert len(boundaries) == 3
|
|
|
|
# test comps2boundaries with blank inputs
|
|
boundaries = comps2boundaries(text_comps[[]], comp_labels[[]])
|
|
assert len(boundaries) == 0
|
|
|
|
|
|
def test_textsnake_decode():
|
|
|
|
maps = torch.zeros((1, 6, 224, 224), dtype=torch.float)
|
|
maps[:, 0:2, :, :] = -10.
|
|
maps[:, 0, 60:100, 50:170] = 10.
|
|
maps[:, 1, 75:85, 60:160] = 10.
|
|
maps[:, 2, 75:85, 60:160] = 0.
|
|
maps[:, 3, 75:85, 60:160] = 1.
|
|
maps[:, 4, 75:85, 60:160] = 10.
|
|
# test decoding with text center region of small area
|
|
maps[:, 0:2, 150:152, 5:7] = 10.
|
|
textsnake_decode = TextSnakePostprocessor()
|
|
results = textsnake_decode(torch.squeeze(maps))
|
|
assert len(results) == 1
|
|
|
|
# test decoding with small radius
|
|
maps.fill_(0.)
|
|
maps[:, 0:2, :, :] = -10.
|
|
maps[:, 0, 120:140, 20:40] = 10.
|
|
maps[:, 1, 120:140, 20:40] = 10.
|
|
maps[:, 2, 120:140, 20:40] = 0.
|
|
maps[:, 3, 120:140, 20:40] = 1.
|
|
maps[:, 4, 120:140, 20:40] = 0.5
|
|
|
|
results = textsnake_decode(torch.squeeze(maps))
|
|
assert len(results) == 0
|
|
|
|
|
|
def test_db_decode():
|
|
pred = torch.zeros((1, 8, 8))
|
|
pred[0, 2:7, 2:7] = 0.8
|
|
expect_result_quad = [[
|
|
1.0, 8.0, 1.0, 1.0, 8.0, 1.0, 8.0, 8.0, 0.800000011920929
|
|
]]
|
|
expect_result_poly = [[
|
|
8, 2, 8, 6, 6, 8, 2, 8, 1, 6, 1, 2, 2, 1, 6, 1, 0.800000011920929
|
|
]]
|
|
with pytest.raises(AssertionError):
|
|
DBPostprocessor(text_repr_type='dummpy')
|
|
db_decode = DBPostprocessor(text_repr_type='quad', min_text_width=1)
|
|
result_quad = db_decode(preds=pred)
|
|
db_decode = DBPostprocessor(text_repr_type='poly', min_text_width=1)
|
|
result_poly = db_decode(preds=pred)
|
|
assert result_quad == expect_result_quad
|
|
assert result_poly == expect_result_poly
|