qurator-spk/eynollah
1
0
Fork 0
mirror of https://github.com/qurator-spk/eynollah.git synced 2026-05-13 01:13:54 +02:00
Code Issues Projects Releases Wiki Activity

resize_image_with_column_classifier(): apply num_col bounds here too

Browse source 
use rules from `resize_and_enhance_image_with_column_classifier()`
and apply them to `resize_image_with_column_classifier()` as well

(to be used by enhancer CLI)
This commit is contained in:
Robert Sachunsky 2026-05-09 02:53:04 +02:00
parent 6df2144c0f
commit d63ce5538c
1 changed files with 31 additions and 12 deletions
Download patch file Download diff file Expand all files Collapse all files

43
src/eynollah/eynollah.py
View file

@ -309,6 +309,7 @@ class Eynollah:
return img_new, img_is_resized
# FIXME: actually may run enhancement model, should be renamed
def resize_image_with_column_classifier(self, image):
self.logger.debug("enter resize_image_with_column_classifier")
img = self.imread(image, binary=self.input_binary)
@ -316,19 +317,36 @@ class Eynollah:
width_early = img.shape[1]
page_img, page_coord = self.early_page_for_num_of_column_classification(img)
if self.input_binary:
img_in = page_img
else:
img_1ch = self.imread(image, grayscale=True, uint8=False)
img_1ch = img_1ch[page_coord[0]: page_coord[1],
page_coord[2]: page_coord[3]]
img_in = np.repeat(img_1ch[:, :, np.newaxis], 3, axis=2)
img_in = img_in / 255.0
img_in = cv2.resize(img_in, (448, 448), interpolation=cv2.INTER_NEAREST).astype(np.float16)
label_p_pred = np.ones(6)
conf_col = 1.0
if self.num_col_upper and not self.num_col_lower:
num_col = self.num_col_upper
elif self.num_col_lower and not self.num_col_upper:
num_col = self.num_col_lower
elif (not self.num_col_upper and not self.num_col_lower or
self.num_col_upper != self.num_col_lower):
if self.input_binary:
img_in = page_img
else:
img_1ch = self.imread(image, grayscale=True)
img_1ch = img_1ch[page_coord[0]: page_coord[1],
page_coord[2]: page_coord[3]]
img_in = np.repeat(img_1ch[:, :, np.newaxis], 3, axis=2)
img_in = img_in / 255.0
img_in = cv2.resize(img_in, (448, 448), interpolation=cv2.INTER_NEAREST).astype(np.float16)
label_p_pred = self.model_zoo.get("col_classifier").predict(img_in[np.newaxis], verbose=0)[0]
num_col = np.argmax(label_p_pred) + 1
conf_col = np.max(label_p_pred)
label_p_pred = self.model_zoo.get("col_classifier").predict(img_in[np.newaxis], verbose=0)[0]
num_col = np.argmax(label_p_pred) + 1
conf_col = np.max(label_p_pred)
if self.num_col_upper and self.num_col_upper < num_col:
num_col = self.num_col_upper
conf_col = 1.0
if self.num_col_lower and self.num_col_lower > num_col:
num_col = self.num_col_lower
conf_col = 1.0
else:
num_col = self.num_col_upper
conf_col = 1.0
self.logger.info("Found %s columns (%s)", num_col, np.around(label_p_pred, decimals=5))
if num_col in (1, 2):
@ -349,6 +367,7 @@ class Eynollah:
image['scale_x'] = 1.0 * img_new.shape[1] / img.shape[1]
return
# FIXME: does not actually run enhancement model, should be renamed
def resize_and_enhance_image_with_column_classifier(self, image):
self.logger.debug("enter resize_and_enhance_image_with_column_classifier")
dpi = image['dpi']