개정판 7c7f8619
issue #506: short text
Change-Id: I7f7b65252ed855621f0fdf9b6644ee81d6d095f7
| DTI_PID/DTI_PID/MainWindow.py | ||
|---|---|---|
| 1570 | 1570 |
def onViewText(self, isChecked): |
| 1571 | 1571 |
selected = [item for item in self.graphicsView.scene().items() if issubclass(type(item), QEngineeringTextItem)] |
| 1572 | 1572 |
for item in selected: |
| 1573 |
item.setVisible(isChecked) |
|
| 1573 |
if type(item) is not QEngineeringLineNoTextItem: |
|
| 1574 |
item.setVisible(isChecked) |
|
| 1574 | 1575 |
|
| 1575 | 1576 |
''' |
| 1576 | 1577 |
@brief visible/invisible Symbol |
| DTI_PID/DTI_PID/OcrResultDialog.py | ||
|---|---|---|
| 363 | 363 |
img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY) |
| 364 | 364 |
imgNot = np.ones(img.shape, np.uint8) |
| 365 | 365 |
cv2.bitwise_not(img, imgNot) |
| 366 |
imgNot = cv2.dilate(imgNot, np.ones((8, 8), np.uint8))
|
|
| 366 |
imgNot = cv2.dilate(imgNot, np.ones((4, 4), np.uint8))
|
|
| 367 | 367 |
|
| 368 | 368 |
contours, hierarchy = cv2.findContours(imgNot, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE) |
| 369 | 369 |
minX, minY, maxX, maxY = sys.maxsize, sys.maxsize, 0, 0 |
| DTI_PID/DTI_PID/tesseract_ocr_module.py | ||
|---|---|---|
| 106 | 106 |
im = Image.fromarray(img) |
| 107 | 107 |
# add padding to increase text recognition |
| 108 | 108 |
im = im.rotate(-angle, expand=True) |
| 109 |
imgWidth = im.width |
|
| 110 |
imgHeight = im.height |
|
| 111 | 109 |
|
| 112 |
#long_img = None |
|
| 113 |
#if imgHeight / imgWidth > 0.5 and imgHeight < maxSize: |
|
| 114 |
# long_img = Image.new('RGB', (imgWidth * 3, imgHeight), (255, 255, 255))
|
|
| 115 |
# long_img.paste(im, (0, 0)) |
|
| 116 |
# long_img.paste(im, (imgWidth, 0)) |
|
| 117 |
# long_img.paste(im, (imgWidth * 2, 0)) |
|
| 118 |
# im = long_img |
|
| 119 |
|
|
| 120 | 110 |
im = ImageOps.expand(im, border=thickness, fill='white') |
| 121 | 111 |
imgWidth = im.width |
| 122 | 112 |
imgHeight = im.height |
| 123 | 113 |
|
| 124 | 114 |
boundaryOcrData = pytesseract.image_to_boxes(im, config=_conf, lang=oCRLang) |
| 125 |
|
|
| 126 | 115 |
bounding_boxes = boundaryOcrData.split('\n')
|
| 116 |
|
|
| 117 |
long_img = None |
|
| 118 |
if not boundaryOcrData and imgHeight / imgWidth > 0.5 and abs(angle) < 1: |
|
| 119 |
img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY) |
|
| 120 |
imgNot = np.ones(img.shape, np.uint8) |
|
| 121 |
cv2.bitwise_not(img, imgNot) |
|
| 122 |
imgNot = cv2.dilate(imgNot, np.ones((3, 3), np.uint8)) |
|
| 123 |
|
|
| 124 |
contours, hierarchy = cv2.findContours(imgNot, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE) |
|
| 125 |
minX, minY, maxX, maxY = sys.maxsize, sys.maxsize, 0, 0 |
|
| 126 |
if len(contours) is 0: |
|
| 127 |
minX, minY, maxX, maxY = 0, 0, imgWidth, imgHeight |
|
| 128 |
else: |
|
| 129 |
minX, minY, maxX, maxY = sys.maxsize, sys.maxsize, 0, 0 |
|
| 130 |
for cnt in contours: |
|
| 131 |
x, y, w, h = cv2.boundingRect(cnt) |
|
| 132 |
minX = min(x, minX) |
|
| 133 |
minY = min(y, minY) |
|
| 134 |
maxX = max(x + w, maxX) |
|
| 135 |
maxY = max(y + h, maxY) |
|
| 136 |
minX, minY, maxX, maxY = minX, minY, maxX, maxY |
|
| 137 |
|
|
| 138 |
fit_width = maxX - minX |
|
| 139 |
fit_length = maxY - minY |
|
| 140 |
if fit_length < maxSize: |
|
| 141 |
long_img = Image.new('RGB', (maxX * 3, maxY), (255, 255, 255))
|
|
| 142 |
fit_im = im.crop((minX + thickness, minY + thickness, minX + thickness + fit_width, minY + thickness + fit_length)) |
|
| 143 |
long_img.paste(fit_im, (0, 0)) |
|
| 144 |
long_img.paste(fit_im, (fit_width, 0)) |
|
| 145 |
long_img.paste(fit_im, (fit_width * 2, 0)) |
|
| 146 |
im = long_img |
|
| 147 |
|
|
| 148 |
im = ImageOps.expand(im, border=thickness, fill='white') |
|
| 149 |
imgWidth = im.width |
|
| 150 |
imgHeight = im.height |
|
| 151 |
offset_x, offset_y = minX, minY |
|
| 152 |
|
|
| 153 |
boundaryOcrData = pytesseract.image_to_boxes(im, config=_conf, lang=oCRLang) |
|
| 154 |
bounding_boxes = boundaryOcrData.split('\n')
|
|
| 155 |
|
|
| 127 | 156 |
merged_boxes = [] |
| 128 | 157 |
|
| 129 | 158 |
for box in bounding_boxes: |
| ... | ... | |
| 168 | 197 |
if cropped.width == 0 or cropped.height == 0: |
| 169 | 198 |
continue |
| 170 | 199 |
text = pytesseract.image_to_string(cropped, config=_conf, lang=oCRLang) |
| 171 |
#if long_img: |
|
| 172 |
# text = text[:int(len(text)/ 3)] |
|
| 173 |
# rect = QRect(rect.left(), imgHeight - rect.bottom(), int((rect.width() - thickness)/ 3), rect.height()) |
|
| 200 |
|
|
| 201 |
if long_img: |
|
| 202 |
text = text[:int(len(text)/ 3)] |
|
| 203 |
rect = QRect(rect.left() + offset_x, imgHeight - rect.bottom() + offset_y, int(rect.width()/ 3), rect.height()) |
|
| 174 | 204 |
|
| 175 | 205 |
# skip if text is not allowed single character when text is one char |
| 176 | 206 |
if not text or (len(text) == 1 and text not in allowed_single_chars): |
| minorTools/makeClearSql.py | ||
|---|---|---|
| 2 | 2 |
import os |
| 3 | 3 |
|
| 4 | 4 |
sqlPath = os.path.join('E:\Projects\DTIPID_GERRIT\minorTools', 'clear.txt')
|
| 5 |
|
|
| 5 |
|
|
| 6 |
drawing_uids = ['1e40e611-cd83-4c99-9711-b45321c3641e', '98240e9a-8ac9-4968-8203-7db2ace5023a'] |
|
| 7 |
|
|
| 8 |
''' |
|
| 6 | 9 |
drawing_uids = ['ca0423e7-bde1-4b64-9d4a-45afa09d2b19', 'cf1166cd-b027-4d5e-88b2-d817e8d9638f', \ |
| 7 | 10 |
'd17b490e-9c33-4451-a727-0cfc076ee8f3', 'd7dd4f37-db5f-433a-b76a-9843027f4307', 'dc4df203-9fd7-47c0-a1c7-704be8f68719', \ |
| 8 | 11 |
'ecbe6ada-dd70-419f-8bf1-0df8edb524f0', 'f1cdd4b3-d54a-48e8-9855-85cdf426142e', 'f52d3484-f6bc-4a4f-b8a8-04f867b4cbb6', \ |
| ... | ... | |
| 21 | 24 |
'18fcfb0d-1529-4cdf-a94f-47edab607bc2', '12bb5341-2038-44e7-b8d7-1d55be426bb2', '10c7e8e1-dd25-47f2-b691-71192693a12d', \ |
| 22 | 25 |
'0a4f828f-68c1-486c-83b9-4be49a443135', '07a5abbd-88d5-460f-89bc-7bb7cedc9746', '0547c696-1963-44c2-9eba-5100297edd40', \ |
| 23 | 26 |
'046e4093-9d9c-49f2-bd7c-af9afedf34d3', '03347731-0f96-4be7-b4bd-223d70c67f75', '016da493-0e9d-42d3-b942-5a45cb36335e'] |
| 27 |
''' |
|
| 24 | 28 |
sqls = [] |
| 25 | 29 |
|
| 26 | 30 |
for index in range(len(drawing_uids)): |
| minorTools/xmlTextSearcher.py | ||
|---|---|---|
| 3 | 3 |
def finder(): |
| 4 | 4 |
print("finder start")
|
| 5 | 5 |
|
| 6 |
targetText = ['t strainer<']
|
|
| 6 |
targetText = ['DESUPERHEATER<']
|
|
| 7 | 7 |
path = 'W:\ID2_Project\POSCO\Temp' |
| 8 | 8 |
|
| 9 | 9 |
print("target : " + str(len(targetText)))
|
내보내기 Unified diff
