Spaces:
Running
Running
Delete app.py
Browse files
app.py
DELETED
@@ -1,292 +0,0 @@
|
|
1 |
-
# -*- encoding: utf-8 -*-
|
2 |
-
# @Author: SWHL
|
3 |
-
# @Contact: [email protected]
|
4 |
-
from enum import Enum
|
5 |
-
from pathlib import Path
|
6 |
-
from typing import List, Union
|
7 |
-
|
8 |
-
import gradio as gr
|
9 |
-
import numpy as np
|
10 |
-
from rapidocr import RapidOCR
|
11 |
-
|
12 |
-
|
13 |
-
class InferEngine(Enum):
|
14 |
-
ort = "ONNXRuntime"
|
15 |
-
vino = "OpenVino"
|
16 |
-
paddle = "PaddlePaddle"
|
17 |
-
torch = "PyTorch"
|
18 |
-
|
19 |
-
|
20 |
-
def get_ocr_engine(infer_engine: str, lang_det: str, lang_rec: str) -> RapidOCR:
|
21 |
-
if infer_engine == InferEngine.vino.value:
|
22 |
-
return RapidOCR(
|
23 |
-
params={
|
24 |
-
"Global.with_openvino": True,
|
25 |
-
"Global.lang_det": lang_det,
|
26 |
-
"Global.lang_rec": lang_rec,
|
27 |
-
}
|
28 |
-
)
|
29 |
-
|
30 |
-
if infer_engine == InferEngine.paddle.value:
|
31 |
-
return RapidOCR(
|
32 |
-
params={
|
33 |
-
"Global.with_paddle": True,
|
34 |
-
"Global.lang_det": lang_det,
|
35 |
-
"Global.lang_rec": lang_rec,
|
36 |
-
}
|
37 |
-
)
|
38 |
-
|
39 |
-
if infer_engine == InferEngine.torch.value:
|
40 |
-
return RapidOCR(
|
41 |
-
params={
|
42 |
-
"Global.with_torch": True,
|
43 |
-
"Global.lang_det": lang_det,
|
44 |
-
"Global.lang_rec": lang_rec,
|
45 |
-
}
|
46 |
-
)
|
47 |
-
|
48 |
-
return RapidOCR(
|
49 |
-
params={
|
50 |
-
"Global.with_onnx": True,
|
51 |
-
"Global.lang_det": lang_det,
|
52 |
-
"Global.lang_rec": lang_rec,
|
53 |
-
}
|
54 |
-
)
|
55 |
-
|
56 |
-
|
57 |
-
def get_ocr_result(
|
58 |
-
img: np.ndarray,
|
59 |
-
text_score,
|
60 |
-
box_thresh,
|
61 |
-
unclip_ratio,
|
62 |
-
lang_det,
|
63 |
-
lang_rec,
|
64 |
-
infer_engine,
|
65 |
-
is_word: str,
|
66 |
-
):
|
67 |
-
return_word_box = True if is_word == "Yes" else False
|
68 |
-
|
69 |
-
ocr_engine = get_ocr_engine(infer_engine, lang_det=lang_det, lang_rec=lang_rec)
|
70 |
-
|
71 |
-
ocr_result = ocr_engine(
|
72 |
-
img,
|
73 |
-
text_score=text_score,
|
74 |
-
box_thresh=box_thresh,
|
75 |
-
unclip_ratio=unclip_ratio,
|
76 |
-
return_word_box=return_word_box,
|
77 |
-
)
|
78 |
-
vis_img = ocr_result.vis()
|
79 |
-
if return_word_box:
|
80 |
-
txts, scores, _ = list(zip(*ocr_result.word_results))
|
81 |
-
ocr_txts = [[i, txt, score] for i, (txt, score) in enumerate(zip(txts, scores))]
|
82 |
-
return vis_img, ocr_txts, ocr_result.elapse
|
83 |
-
|
84 |
-
ocr_txts = [
|
85 |
-
[i, txt, score]
|
86 |
-
for i, (txt, score) in enumerate(zip(ocr_result.txts, ocr_result.scores))
|
87 |
-
]
|
88 |
-
return vis_img, ocr_txts, ocr_result.elapse
|
89 |
-
|
90 |
-
|
91 |
-
def create_examples() -> List[List[Union[str, float]]]:
|
92 |
-
examples = [
|
93 |
-
[
|
94 |
-
"images/ch_en_num.jpg",
|
95 |
-
0.5,
|
96 |
-
0.5,
|
97 |
-
1.6,
|
98 |
-
"ch_mobile",
|
99 |
-
"ch_mobile",
|
100 |
-
"ONNXRuntime",
|
101 |
-
"No",
|
102 |
-
],
|
103 |
-
[
|
104 |
-
"images/japan.jpg",
|
105 |
-
0.5,
|
106 |
-
0.5,
|
107 |
-
1.6,
|
108 |
-
"multi_mobile",
|
109 |
-
"japan_mobile",
|
110 |
-
"ONNXRuntime",
|
111 |
-
"No",
|
112 |
-
],
|
113 |
-
[
|
114 |
-
"images/korean.jpg",
|
115 |
-
0.5,
|
116 |
-
0.5,
|
117 |
-
1.6,
|
118 |
-
"multi_mobile",
|
119 |
-
"korean_mobile",
|
120 |
-
"ONNXRuntime",
|
121 |
-
"No",
|
122 |
-
],
|
123 |
-
[
|
124 |
-
"images/air_ticket.jpg",
|
125 |
-
0.5,
|
126 |
-
0.5,
|
127 |
-
1.6,
|
128 |
-
"ch_mobile",
|
129 |
-
"ch_mobile",
|
130 |
-
"ONNXRuntime",
|
131 |
-
"No",
|
132 |
-
],
|
133 |
-
[
|
134 |
-
"images/car_plate.jpeg",
|
135 |
-
0.5,
|
136 |
-
0.5,
|
137 |
-
1.6,
|
138 |
-
"ch_mobile",
|
139 |
-
"ch_mobile",
|
140 |
-
"ONNXRuntime",
|
141 |
-
"No",
|
142 |
-
],
|
143 |
-
[
|
144 |
-
"images/train_ticket.jpeg",
|
145 |
-
0.5,
|
146 |
-
0.5,
|
147 |
-
1.6,
|
148 |
-
"ch_mobile",
|
149 |
-
"ch_mobile",
|
150 |
-
"ONNXRuntime",
|
151 |
-
"No",
|
152 |
-
],
|
153 |
-
]
|
154 |
-
return examples
|
155 |
-
|
156 |
-
|
157 |
-
infer_engine_list = [InferEngine[v].value for v in InferEngine.__members__]
|
158 |
-
|
159 |
-
lang_det_list = ["ch_mobile", "ch_server", "en_mobile", "en_server", "multi_mobile"]
|
160 |
-
lang_rec_list = [
|
161 |
-
"ch_mobile",
|
162 |
-
"ch_server",
|
163 |
-
"chinese_cht",
|
164 |
-
"en_mobile",
|
165 |
-
"ar_mobile",
|
166 |
-
"cyrillic_mobile",
|
167 |
-
"devanagari_mobile",
|
168 |
-
"japan_mobile",
|
169 |
-
"ka_mobile",
|
170 |
-
"korean_mobile",
|
171 |
-
"latin_mobile",
|
172 |
-
"ta_mobile",
|
173 |
-
"te_mobile",
|
174 |
-
]
|
175 |
-
|
176 |
-
custom_css = """
|
177 |
-
body {font-family: body {font-family: 'Helvetica Neue', Helvetica;}
|
178 |
-
.gr-button {background-color: #4CAF50; color: white; border: none; padding: 10px 20px; border-radius: 5px;}
|
179 |
-
.gr-button:hover {background-color: #45a049;}
|
180 |
-
.gr-textbox {margin-bottom: 15px;}
|
181 |
-
.example-button {background-color: #1E90FF; color: white; border: none; padding: 8px 15px; border-radius: 5px; margin: 5px;}
|
182 |
-
.example-button:hover {background-color: #FF4500;}
|
183 |
-
.tall-radio .gr-radio-item {padding: 15px 0; min-height: 50px; display: flex; align-items: center;}
|
184 |
-
.tall-radio label {font-size: 16px;}
|
185 |
-
.output-image, .input-image, .image-preview {height: 300px !important}
|
186 |
-
"""
|
187 |
-
|
188 |
-
with gr.Blocks(
|
189 |
-
title="Rapid⚡OCR Demo", css="custom_css", theme=gr.themes.Soft()
|
190 |
-
) as demo:
|
191 |
-
gr.Markdown(
|
192 |
-
"<h1 style='text-align: center;'><a href='https://rapidai.github.io/RapidOCRDocs/' style='text-decoration: none;'>Rapid⚡OCR</a></h1>"
|
193 |
-
)
|
194 |
-
gr.HTML(
|
195 |
-
"""
|
196 |
-
<div style="display: flex; justify-content: center; gap: 10px;">
|
197 |
-
<a href=""><img src="https://img.shields.io/badge/Python->=3.6-aff.svg"></a>
|
198 |
-
<a href=""><img src="https://img.shields.io/badge/OS-Linux%2C%20Win%2C%20Mac-pink.svg"></a>
|
199 |
-
<a href="https://pepy.tech/project/rapidocr"><img src="https://static.pepy.tech/personalized-badge/rapidocr?period=total&units=abbreviation&left_color=grey&right_color=blue&left_text=Downloads%20rapidocr"></a>
|
200 |
-
<a href="https://pypi.org/project/rapidocr/"><img alt="PyPI" src="https://img.shields.io/pypi/v/rapidocr"></a>
|
201 |
-
<a href="https://github.com/RapidAI/RapidOCR"><img src="https://img.shields.io/github/stars/RapidAI/RapidOCR?color=ccf"></a>
|
202 |
-
</div>
|
203 |
-
"""
|
204 |
-
)
|
205 |
-
with gr.Row():
|
206 |
-
text_score = gr.Slider(
|
207 |
-
label="text_score",
|
208 |
-
minimum=0,
|
209 |
-
maximum=1.0,
|
210 |
-
value=0.5,
|
211 |
-
step=0.1,
|
212 |
-
info="文本识别结果是正确的置信度,值越大,显示出的识别结果更准确。存在漏检时,调低该值。取值范围:[0, 1.0],默认值为0.5",
|
213 |
-
)
|
214 |
-
box_thresh = gr.Slider(
|
215 |
-
label="box_thresh",
|
216 |
-
minimum=0,
|
217 |
-
maximum=1.0,
|
218 |
-
value=0.5,
|
219 |
-
step=0.1,
|
220 |
-
info="检测到的框是文本的概率,值越大,框中是文本的概率就越大。存在漏检时,调低该值。取值范围:[0, 1.0],默认值为0.5",
|
221 |
-
)
|
222 |
-
unclip_ratio = gr.Slider(
|
223 |
-
label="unclip_ratio",
|
224 |
-
minimum=1.5,
|
225 |
-
maximum=2.0,
|
226 |
-
value=1.6,
|
227 |
-
step=0.1,
|
228 |
-
info="控制文本检测框的大小,值越大,检测框整体越大。在出现框截断文字的情况,调大该值。取值范围:[1.5, 2.0],默认值为1.6",
|
229 |
-
)
|
230 |
-
|
231 |
-
with gr.Row():
|
232 |
-
select_infer_engine = gr.Dropdown(
|
233 |
-
choices=infer_engine_list,
|
234 |
-
label="Infer Engine (推理引擎)",
|
235 |
-
value="ONNXRuntime",
|
236 |
-
interactive=True,
|
237 |
-
)
|
238 |
-
lang_det = gr.Dropdown(
|
239 |
-
choices=lang_det_list,
|
240 |
-
label="Det model (文本检测模型)",
|
241 |
-
value=lang_det_list[0],
|
242 |
-
interactive=True,
|
243 |
-
)
|
244 |
-
lang_rec = gr.Dropdown(
|
245 |
-
choices=lang_rec_list,
|
246 |
-
label="Rec model (文本识别模型)",
|
247 |
-
value=lang_rec_list[0],
|
248 |
-
interactive=True,
|
249 |
-
)
|
250 |
-
is_word = gr.Radio(
|
251 |
-
["Yes", "No"], label="Return word box (返回单字符)", value="No"
|
252 |
-
)
|
253 |
-
|
254 |
-
img_input = gr.Image(label="Upload or Select Image", sources="upload")
|
255 |
-
|
256 |
-
run_btn = gr.Button("Run")
|
257 |
-
|
258 |
-
img_output = gr.Image(label="Output Image")
|
259 |
-
elapse = gr.Textbox(label="Elapse(s)")
|
260 |
-
ocr_results = gr.Dataframe(
|
261 |
-
label="OCR Txts",
|
262 |
-
headers=["Index", "Txt", "Score"],
|
263 |
-
datatype=["number", "str", "number"],
|
264 |
-
show_copy_button=True,
|
265 |
-
)
|
266 |
-
|
267 |
-
ocr_inputs = [
|
268 |
-
img_input,
|
269 |
-
text_score,
|
270 |
-
box_thresh,
|
271 |
-
unclip_ratio,
|
272 |
-
lang_det,
|
273 |
-
lang_rec,
|
274 |
-
select_infer_engine,
|
275 |
-
is_word,
|
276 |
-
]
|
277 |
-
run_btn.click(
|
278 |
-
get_ocr_result, inputs=ocr_inputs, outputs=[img_output, ocr_results, elapse]
|
279 |
-
)
|
280 |
-
|
281 |
-
examples = gr.Examples(
|
282 |
-
examples=create_examples(),
|
283 |
-
examples_per_page=5,
|
284 |
-
inputs=ocr_inputs,
|
285 |
-
fn=get_ocr_result,
|
286 |
-
outputs=[img_output, ocr_results, elapse],
|
287 |
-
cache_examples=False,
|
288 |
-
)
|
289 |
-
|
290 |
-
|
291 |
-
if __name__ == "__main__":
|
292 |
-
demo.launch(debug=True)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|