|
|
import numpy as np |
|
|
import cv2 |
|
|
from typing import Tuple, List |
|
|
|
|
|
import sys,os |
|
|
currDir = os.path.dirname(os.path.abspath(__file__)) |
|
|
rootDir = os.path.dirname( os.path.dirname(currDir) ) |
|
|
sys.path.append(rootDir) |
|
|
|
|
|
if __name__ == "__main__": |
|
|
from base import register_textdetectors, TextDetectorBase, TextBlock, DEFAULT_DEVICE, DEVICE_SELECTOR, ProjImgTrans |
|
|
from ctd import CTDModel |
|
|
else: |
|
|
from .base import register_textdetectors, TextDetectorBase, TextBlock, DEFAULT_DEVICE, DEVICE_SELECTOR, ProjImgTrans |
|
|
from .ctd import CTDModel |
|
|
|
|
|
CTD_ONNX_PATH = 'data/models/comictextdetector.pt.onnx' |
|
|
CTD_TORCH_PATH = 'data/models/comictextdetector.pt' |
|
|
|
|
|
def load_ctd_model(model_path, device, detect_size=1024) -> CTDModel: |
|
|
model = CTDModel(model_path, detect_size=detect_size, device=device) |
|
|
|
|
|
return model |
|
|
|
|
|
@register_textdetectors('ctd') |
|
|
class ComicTextDetector(TextDetectorBase): |
|
|
|
|
|
params = { |
|
|
'detect_size': { |
|
|
'type': 'selector', |
|
|
'options': [896, 1024, 1152, 1280], |
|
|
'value': 1280 |
|
|
}, |
|
|
'det_rearrange_max_batches': { |
|
|
'type': 'selector', |
|
|
'options': [1, 2, 4, 6, 8, 12, 16, 24, 32], |
|
|
'value': 4 |
|
|
}, |
|
|
'device': DEVICE_SELECTOR(), |
|
|
'description': 'ComicTextDetector', |
|
|
'font size multiplier': 1., |
|
|
'font size max': -1, |
|
|
'font size min': -1, |
|
|
'mask dilate size': 2 |
|
|
} |
|
|
_load_model_keys = {'model'} |
|
|
download_file_list = [{ |
|
|
'url': 'https://github.com/zyddnys/manga-image-translator/releases/download/beta-0.3/', |
|
|
'files': ['data/models/comictextdetector.pt', 'data/models/comictextdetector.pt.onnx'], |
|
|
'sha256_pre_calculated': ['1f90fa60aeeb1eb82e2ac1167a66bf139a8a61b8780acd351ead55268540cccb', '1a86ace74961413cbd650002e7bb4dcec4980ffa21b2f19b86933372071d718f'], |
|
|
'concatenate_url_filename': 2, |
|
|
}] |
|
|
|
|
|
device = DEFAULT_DEVICE |
|
|
detect_size = 1024 |
|
|
def __init__(self, **params) -> None: |
|
|
super().__init__(**params) |
|
|
self.model: CTDModel = None |
|
|
|
|
|
@property |
|
|
def device(self): |
|
|
return self.params['device']['value'] |
|
|
|
|
|
@property |
|
|
def detect_size(self): |
|
|
return int(self.params['detect_size']['value']) |
|
|
|
|
|
def _load_model(self): |
|
|
if self.device != 'cpu': |
|
|
self.model = load_ctd_model(CTD_TORCH_PATH, self.device, self.detect_size) |
|
|
else: |
|
|
self.model = load_ctd_model(CTD_ONNX_PATH, self.device, self.detect_size) |
|
|
|
|
|
def _detect(self, img: np.ndarray, proj: ProjImgTrans) -> Tuple[np.ndarray, List[TextBlock]]: |
|
|
_, mask, blk_list = self.model(img) |
|
|
|
|
|
fnt_rsz = self.get_param_value('font size multiplier') |
|
|
fnt_max = self.get_param_value('font size max') |
|
|
fnt_min = self.get_param_value('font size min') |
|
|
for blk in blk_list: |
|
|
sz = blk._detected_font_size * fnt_rsz |
|
|
if fnt_max > 0: |
|
|
sz = min(fnt_max, sz) |
|
|
if fnt_min > 0: |
|
|
sz = max(fnt_min, sz) |
|
|
blk.font_size = sz |
|
|
blk._detected_font_size = sz |
|
|
|
|
|
ksize = self.get_param_value('mask dilate size') |
|
|
if ksize > 0: |
|
|
element = cv2.getStructuringElement(cv2.MORPH_ELLIPSE, (2 * ksize + 1, 2 * ksize + 1),(ksize, ksize)) |
|
|
mask = cv2.dilate(mask, element) |
|
|
|
|
|
return mask, blk_list |
|
|
|
|
|
def updateParam(self, param_key: str, param_content): |
|
|
super().updateParam(param_key, param_content) |
|
|
device = self.device |
|
|
if self.model is not None: |
|
|
if self.model.device != device: |
|
|
self.model.device = device |
|
|
if device != 'cpu': |
|
|
self.model.load_model(CTD_TORCH_PATH) |
|
|
else: |
|
|
self.model.load_model(CTD_ONNX_PATH) |
|
|
self.model.detect_size = self.detect_size |
|
|
|
|
|
|
|
|
if __name__ == '__main__': |
|
|
model = load_ctd_model(CTD_ONNX_PATH, 'cpu', 1024) |
|
|
pass |
