Skip to content

A
accessibility_movie_2
Commit 8142abd4 authored by 翟艳秋（20软）'s avatar 翟艳秋（20软）
Browse files
Options

fix some bugs and fit ocr with low quality films

parent ad04a182
Hide whitespace changes
Inline Side-by-side
Showing with 600 additions and 656 deletions
assemble_dialog.ui
View file @ 8142abd4
......@@ -13,155 +13,72 @@
<property name="windowTitle">
<string>Dialog</string>
</property>
<widget class="QDialogButtonBox" name="buttonBox">
<property name="geometry">
<rect>
<x>220</x>
<y>250</y>
<width>341</width>
<height>32</height>
</rect>
</property>
<property name="toolTip">
<string/>
</property>
<property name="toolTipDuration">
<number>-1</number>
</property>
<property name="orientation">
<enum>Qt::Horizontal</enum>
</property>
<property name="standardButtons">
<set>QDialogButtonBox::Cancel|QDialogButtonBox::Ok</set>
</property>
<property name="centerButtons">
<bool>false</bool>
</property>
</widget>
<widget class="QLabel" name="label">
<property name="geometry">
<rect>
<x>40</x>
<y>40</y>
<width>72</width>
<height>15</height>
</rect>
</property>
<property name="text">
<string>视频路径:</string>
</property>
</widget>
<widget class="QLineEdit" name="lineEdit">
<property name="geometry">
<rect>
<x>130</x>
<y>40</y>
<width>231</width>
<height>21</height>
</rect>
</property>
</widget>
<widget class="QPushButton" name="pushButton">
<property name="geometry">
<rect>
<x>390</x>
<y>40</y>
<width>101</width>
<height>31</height>
</rect>
</property>
<property name="text">
<string>选择文件</string>
</property>
</widget>
<widget class="QLabel" name="label_2">
<property name="geometry">
<rect>
<x>20</x>
<y>90</y>
<width>111</width>
<height>16</height>
</rect>
</property>
<property name="text">
<string>旁白脚本表格：</string>
</property>
</widget>
<widget class="QLineEdit" name="lineEdit_2">
<property name="geometry">
<rect>
<x>130</x>
<y>90</y>
<width>231</width>
<height>21</height>
</rect>
</property>
</widget>
<widget class="QPushButton" name="pushButton_2">
<property name="geometry">
<rect>
<x>390</x>
<y>90</y>
<width>101</width>
<height>31</height>
</rect>
</property>
<property name="text">
<string>表格路径</string>
</property>
</widget>
<widget class="QLabel" name="label_3">
<property name="geometry">
<rect>
<x>30</x>
<y>140</y>
<width>111</width>
<height>16</height>
</rect>
</property>
<property name="text">
<string>旁白说话人：</string>
</property>
</widget>
<widget class="QLabel" name="label_4">
<property name="geometry">
<rect>
<x>30</x>
<y>200</y>
<width>111</width>
<height>16</height>
</rect>
</property>
<property name="text">
<string>旁白语速：</string>
</property>
</widget>
<widget class="QLineEdit" name="lineEdit_3">
<property name="enabled">
<bool>false</bool>
</property>
<property name="geometry">
<rect>
<x>130</x>
<y>140</y>
<width>231</width>
<height>21</height>
</rect>
</property>
</widget>
<widget class="QLineEdit" name="lineEdit_4">
<property name="enabled">
<bool>false</bool>
</property>
<property name="geometry">
<rect>
<x>130</x>
<y>200</y>
<width>231</width>
<height>21</height>
</rect>
</property>
</widget>
<layout class="QGridLayout" name="gridLayout">
<item row="0" column="2">
<widget class="QPushButton" name="pushButton_2">
<property name="text">
<string>表格路径</string>
</property>
</widget>
</item>
<item row="1" column="1">
<widget class="QLineEdit" name="lineEdit_3">
<property name="enabled">
<bool>false</bool>
</property>
</widget>
</item>
<item row="2" column="1">
<widget class="QLineEdit" name="lineEdit_4">
<property name="enabled">
<bool>false</bool>
</property>
</widget>
</item>
<item row="0" column="1">
<widget class="QLineEdit" name="lineEdit_2"/>
</item>
<item row="2" column="0">
<widget class="QLabel" name="label_4">
<property name="text">
<string>旁白语速：</string>
</property>
</widget>
</item>
<item row="3" column="1" alignment="Qt::AlignHCenter">
<widget class="QDialogButtonBox" name="buttonBox">
<property name="toolTip">
<string/>
</property>
<property name="toolTipDuration">
<number>-1</number>
</property>
<property name="orientation">
<enum>Qt::Horizontal</enum>
</property>
<property name="standardButtons">
<set>QDialogButtonBox::Cancel|QDialogButtonBox::Ok</set>
</property>
<property name="centerButtons">
<bool>false</bool>
</property>
</widget>
</item>
<item row="0" column="0">
<widget class="QLabel" name="label_2">
<property name="text">
<string>旁白脚本表格：</string>
</property>
</widget>
</item>
<item row="1" column="0">
<widget class="QLabel" name="label_3">
<property name="text">
<string>旁白说话人：</string>
</property>
</widget>
</item>
</layout>
</widget>
<resources/>
<connections>
......
assemble_dialog_ui.py
View file @ 8142abd4
# -*- coding: utf-8 -*-
# Form implementation generated from reading ui file 'assemble_dialog.ui'
# Form implementation generated from reading ui file 'd:\AddCaption\cur_version\accessibility_movie_2\assemble_dialog.ui'
#
# Created by: PyQt5 UI code generator 5.12
# Created by: PyQt5 UI code generator 5.15.4
#
# WARNING! All changes made in this file will be lost!
# WARNING: Any manual changes made to this file will be lost when pyuic5 is
# run again. Do not edit this file unless you know what you are doing.
from PyQt5 import QtCore, QtGui, QtWidgets
......@@ -13,46 +15,48 @@ class Ui_Dialog(object):
def setupUi(self, Dialog):
Dialog.setObjectName("Dialog")
Dialog.resize(585, 326)
self.buttonBox = QtWidgets.QDialogButtonBox(Dialog)
self.buttonBox.setGeometry(QtCore.QRect(220, 250, 341, 32))
self.buttonBox.setToolTip("")
self.buttonBox.setToolTipDuration(-1)
self.buttonBox.setOrientation(QtCore.Qt.Horizontal)
self.buttonBox.setStandardButtons(QtWidgets.QDialogButtonBox.Cancel|QtWidgets.QDialogButtonBox.Ok)
self.buttonBox.setCenterButtons(False)
self.buttonBox.setObjectName("buttonBox")
self.label = QtWidgets.QLabel(Dialog)
self.label.setGeometry(QtCore.QRect(40, 40, 72, 15))
self.label.setObjectName("label")
self.lineEdit = QtWidgets.QLineEdit(Dialog)
self.lineEdit.setGeometry(QtCore.QRect(130, 40, 231, 21))
self.lineEdit.setObjectName("lineEdit")
self.gridLayout = QtWidgets.QGridLayout(Dialog)
self.gridLayout.setObjectName("gridLayout")
self.pushButton = QtWidgets.QPushButton(Dialog)
self.pushButton.setGeometry(QtCore.QRect(390, 40, 101, 31))
self.pushButton.setObjectName("pushButton")
self.label_2 = QtWidgets.QLabel(Dialog)
self.label_2.setGeometry(QtCore.QRect(20, 90, 111, 16))
self.label_2.setObjectName("label_2")
self.lineEdit_2 = QtWidgets.QLineEdit(Dialog)
self.lineEdit_2.setGeometry(QtCore.QRect(130, 90, 231, 21))
self.lineEdit_2.setObjectName("lineEdit_2")
self.pushButton_2 = QtWidgets.QPushButton(Dialog)
self.pushButton_2.setGeometry(QtCore.QRect(390, 90, 101, 31))
self.pushButton_2.setObjectName("pushButton_2")
self.label_3 = QtWidgets.QLabel(Dialog)
self.label_3.setGeometry(QtCore.QRect(30, 140, 111, 16))
self.label_3.setObjectName("label_3")
self.gridLayout.addWidget(self.pushButton, 0, 2, 1, 1)
self.label_4 = QtWidgets.QLabel(Dialog)
self.label_4.setGeometry(QtCore.QRect(30, 200, 111, 16))
self.label_4.setObjectName("label_4")
self.gridLayout.addWidget(self.label_4, 3, 0, 1, 1)
self.label_3 = QtWidgets.QLabel(Dialog)
self.label_3.setObjectName("label_3")
self.gridLayout.addWidget(self.label_3, 2, 0, 1, 1)
self.lineEdit_3 = QtWidgets.QLineEdit(Dialog)
self.lineEdit_3.setEnabled(False)
self.lineEdit_3.setGeometry(QtCore.QRect(130, 140, 231, 21))
self.lineEdit_3.setObjectName("lineEdit_3")
self.gridLayout.addWidget(self.lineEdit_3, 2, 1, 1, 1)
self.lineEdit_4 = QtWidgets.QLineEdit(Dialog)
self.lineEdit_4.setEnabled(False)
self.lineEdit_4.setGeometry(QtCore.QRect(130, 200, 231, 21))
self.lineEdit_4.setObjectName("lineEdit_4")
self.gridLayout.addWidget(self.lineEdit_4, 3, 1, 1, 1)
self.lineEdit = QtWidgets.QLineEdit(Dialog)
self.lineEdit.setObjectName("lineEdit")
self.gridLayout.addWidget(self.lineEdit, 0, 1, 1, 1)
self.label_2 = QtWidgets.QLabel(Dialog)
self.label_2.setObjectName("label_2")
self.gridLayout.addWidget(self.label_2, 1, 0, 1, 1)
self.lineEdit_2 = QtWidgets.QLineEdit(Dialog)
self.lineEdit_2.setObjectName("lineEdit_2")
self.gridLayout.addWidget(self.lineEdit_2, 1, 1, 1, 1)
self.label = QtWidgets.QLabel(Dialog)
self.label.setObjectName("label")
self.gridLayout.addWidget(self.label, 0, 0, 1, 1)
self.pushButton_2 = QtWidgets.QPushButton(Dialog)
self.pushButton_2.setObjectName("pushButton_2")
self.gridLayout.addWidget(self.pushButton_2, 1, 2, 1, 1)
self.buttonBox = QtWidgets.QDialogButtonBox(Dialog)
self.buttonBox.setToolTip("")
self.buttonBox.setToolTipDuration(-1)
self.buttonBox.setOrientation(QtCore.Qt.Horizontal)
self.buttonBox.setStandardButtons(QtWidgets.QDialogButtonBox.Cancel|QtWidgets.QDialogButtonBox.Ok)
self.buttonBox.setCenterButtons(False)
self.buttonBox.setObjectName("buttonBox")
self.gridLayout.addWidget(self.buttonBox, 4, 1, 1, 1, QtCore.Qt.AlignHCenter)
self.retranslateUi(Dialog)
self.buttonBox.accepted.connect(Dialog.accept)
......@@ -62,9 +66,9 @@ class Ui_Dialog(object):
def retranslateUi(self, Dialog):
_translate = QtCore.QCoreApplication.translate
Dialog.setWindowTitle(_translate("Dialog", "Dialog"))
self.label.setText(_translate("Dialog", "视频路径:"))
self.pushButton.setText(_translate("Dialog", "选择文件"))
self.label_4.setText(_translate("Dialog", "旁白语速："))
self.label_3.setText(_translate("Dialog", "旁白说话人："))
self.label_2.setText(_translate("Dialog", "旁白脚本表格："))
self.label.setText(_translate("Dialog", "视频路径:"))
self.pushButton_2.setText(_translate("Dialog", "表格路径"))
self.label_3.setText(_translate("Dialog", "旁白说话人："))
self.label_4.setText(_translate("Dialog", "旁白语速："))
\ No newline at end of file
detect_with_ocr.py
View file @ 8142abd4
"""基于OCR的旁白区间检测算法
总体流程如下：
- 首先检测得到字幕的上下边界位置，框定大概的字幕检测范围；
- 视频每秒取3帧，并使用OCR获取字幕以及时间戳等信息；
- 根据字幕间的时间差判断能否插入旁白以及旁白的推荐字数。
外部调用方式：
.. code-block:: python
from detect_with_ocr import detect_with_ocr
detect_with_ocr(video_path, book_path, start_time, end_time, state)
"""
import os
import cv2
import numpy as np
from paddleocr import PaddleOCR
import sys
print("PaddleOCR load path:", os.path.abspath(sys.modules[PaddleOCR.__module__].__file__))
import difflib
import re
from typing import Tuple, Union
from detect_with_asr import create_sheet, write_to_sheet
from main_window import MainWindow, Element
# 字幕的上下边界
up_b, down_b = 0, 0
# 初始化ocr工具
ocr = PaddleOCR(use_angle_cls=True, lang="ch", show_log=False, use_gpu=False)
# 正常语速为4字/秒
normal_speed = 4
def get_position(video_path: str, start_time: float) -> Tuple[float, float]:
# return (885.0, 989.0)
"""根据对视频中的画面进行分析，确定字幕的位置，以便后续的字幕识别
Args:
video_path (str): 视频存储路径
start_time (float): 视频开始时间
Returns:
Tuple[float, float]: 字幕在整个画面中的上下边界位置
"""
print("video_path:", video_path)
video = cv2.VideoCapture(video_path)
# print("video:", video)
subtitle_position = {}
fps = video.get(cv2.CAP_PROP_FPS)
start = int(start_time * fps)
cnt = 0
txt_cnt = 0
pre_txt = None
video.set(cv2.CAP_PROP_POS_FRAMES, start)
height = int(video.get(cv2.CAP_PROP_FRAME_HEIGHT) * 0.6)
while True:
_, img = video.read()
# print("img:", img)
# gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
# cv2.imshow('img', gray)
# cv2.imshow(img)
cnt += 1
if img is None or cnt > 10000:
break
if cnt % int(fps / 3) != 0:
continue
img = img[height:]
res = ocr.ocr(img, cls=True)
sorted(res, key=lambda text: text[0][0][1])
bottom_position = None
if len(res) == 0:
continue
log = []
print("cnt:", cnt, "rect_num:", len(res))
for x in res:
# print("x:", x)
rect, (txt, confidence) = x
# font_size = rect[2][1] - rect[0][1]
mid = (rect[0][0] + rect[1][0]) / 2
gradient = np.arctan(abs((rect[1][1] - rect[0][1]) / (rect[1][0] - rect[0][0])))
# 可能是字幕的文本
if confidence > 0.9 and 0.4 * img.shape[1] < mid < 0.6 * img.shape[1] and gradient < 0.1:
if bottom_position is None:
bottom_position = rect[0][1]
# 判断是否与前一文本相同（是不是同一个字幕），非同一字幕的前提下，取对应上下边界，
keys = subtitle_position.keys()
if abs(rect[0][1] - bottom_position) < 10:
if pre_txt is None or pre_txt != txt:
txt_cnt += 1
pre_txt = txt
if (rect[0][0], rect[2][1]) in keys:
subtitle_position[(rect[0][1], rect[2][1])] += 1
else:
replace = False
for k in keys:
# 更新键值为最宽的上下限
if abs(rect[0][1] - k[0]) + abs(rect[2][1] - k[1]) < 10:
new_k = min(k[0], rect[0][1]), max(k[1], rect[2][1])
if new_k != k:
subtitle_position[new_k] = subtitle_position[k]
subtitle_position[new_k] += 1
subtitle_position.pop(k)
else:
subtitle_position[k] += 1
replace = True
break
if not replace:
subtitle_position[(rect[0][1], rect[2][1])] = 1
if txt_cnt == 3:
break
print(subtitle_position)
up_bounding, down_bounding = max(subtitle_position, key=subtitle_position.get)
return up_bounding + height, down_bounding + height
def erasePunc(txt: str) -> str:
"""去除字符串中的非中文字符
Args:
txt (str): 待处理字符串
Returns:
str: 处理后的字符串
"""
pattern = re.compile(r'[^\u4e00-\u9fa5]')
txt = re.sub(pattern, '', txt)
return txt
def string_similar(s1: str, s2: str) -> float:
"""比较字符串s1和s2的相似度，主要用于减少输出文件中相似字幕的重复
Args:
s1 (str): 第一个字符串
s2 (str): 第二个字符串
Returns:
float: 字符串间的相似度
"""
# 去除非中文字符后，再比较相似度
s1 = erasePunc(s1)
s2 = erasePunc(s2)
return difflib.SequenceMatcher(None, s1, s2).quick_ratio()
def normalize(text: str) -> str:
"""规范化处理文本中的一些标点符号
Args:
text (str): 待处理字符串
Returns:
str: 处理后的字符串
"""
# 将英文标点转换为中文标点
E_pun = u',.!?()[]:;'
C_pun = u'，。！？（）【】：；'
table = {ord(f): ord(t) for f, t in zip(E_pun, C_pun)}
text = text.translate(table)
text = text.strip(' ，。、【】_·：-@‘［；')
# 促成首尾匹配的（）
if text[-1] == '）' and text[0] != '（':
text = '（' + text
elif text[-1] != '）' and text[0] == '（':
text = text + '）'
return text
def detect_subtitle(img: np.ndarray) -> Union[str, None]:
"""检测当前画面得到字幕信息
Args:
img (np.ndarray): 当前画面
Returns:
Union[str, None]: 字幕信息(没有字幕时返回None)
"""
subTitle = ''
height = down_b - up_b
img = img[int(up_b - height * 0.7):int(down_b + height * 0.7)]
# img = cv2.resize(img, (int(img.shape[1] * 0.5), int(img.shape[0] * 0.5)))
res = ocr.ocr(img, cls=True)
sorted(res, key=lambda text: text[0][0][1])
if len(res) == 0:
return None
possible_txt = []
for x in res:
# cv2.imshow("cut", img)
# cv2.waitKey(0)
# cv2.destroyAllWindows()
rect, (txt, confidence) = x
font_size = rect[2][1] - rect[0][1]
mid = (rect[0][0] + rect[1][0]) / 2
gradient = np.arctan(abs((rect[1][1] - rect[0][1]) / (rect[1][0] - rect[0][0])))
# log.append("文本：{}，置信度：{}，中心点：{}，斜率：{}，字体大小：{}".format(txt, confidence, mid / img.shape[1], gradient,
# font_size)) 置信度>0.7 & 斜率<0.1 & 字幕偏移量<=25 & 字幕中心在画面宽的0.4-0.6之间
if confidence > 0.7 and gradient < 0.1 and 0.4 < mid / img.shape[1] < 0.6 and \
abs(rect[0][1] - 30) + abs(img.shape[0] - rect[2][1] - 30) <= 25:
subTitle += txt
# 如果字幕在一行中分为两个（或以上）对话文本
elif confidence > 0.85 and gradient < 0.1:
if 0.3 < mid / img.shape[1] < 0.4 or 0.6 < mid / img.shape[1] < 0.7:
possible_txt.append(txt)
if len(possible_txt) >= 2:
subTitle = ''.join(possible_txt)
if len(subTitle) > 0:
return subTitle
return None
def process_video(video_path: str, begin: float, end: float, book_path: str, sheet_name: str, state=None, mainWindow: MainWindow=None):
"""处理视频，主要完成对字幕的捕捉以及根据字幕分析得出旁白可能位置的任务
Args:
video_path (str): 待处理视频的路径
begin (float): 电影的实际开始位置（秒）
end (float): 电影除演职表外的实际结束位置（秒）
book_path (str): 输出表格地址
sheet_name (str): 输出表格中的表名
state (optional): 用于通信的状态关键字. Defaults to None.
"""
global normal_speed
if mainWindow.projectContext.speaker_speed is not None:
normal_speed = float(mainWindow.projectContext.speaker_speed.split('(')[1].split('字')[0])
sz = len(mainWindow.projectContext.all_elements)
if sz == 0:
last_time = begin
else:
last_time = float(mainWindow.projectContext.all_elements[sz - 1].ed_time_sec) + 0.01
print("当前使用的语速为", normal_speed)
if state is None:
state = [None]
video = cv2.VideoCapture(video_path)
fps = video.get(cv2.CAP_PROP_FPS)
lastSubTitle = None
# res是在视频遍历过程中获取的字幕文件，不掺杂对旁白的分析
res = []
cnt = 0
start_time = 0
end_time = 0
video.set(cv2.CAP_PROP_POS_MSEC, begin * 1000)
pre_state = state[0]
while True:
_, frame = video.read()
if frame is None:
break
cnt += 1
cur_time = video.get(cv2.CAP_PROP_POS_MSEC) / 1000
# 判断当前帧是否已超限制
if cur_time > end:
if cur_time - end_time > 1:
print('--------------------------------------------------')
recommend_lens = int((cur_time - end_time) * normal_speed)
# write_to_sheet(book_path, sheet_name, ['', '', '', '插入旁白，推荐字数为%d' % recommend_lens])
add_to_list(mainWindow, "旁白", ['', '', '', '插入旁白，推荐字数为%d' % recommend_lens])
# 判断当前是否有字幕需要被保存下来
if end_time < start_time:
# write_to_sheet(book_path, sheet_name, [round(start_time, 2), round(end, 2), lastSubTitle, ''])
add_to_list(mainWindow, "字幕", [round(start_time, 3), round(end, 3), lastSubTitle, ''])
break
# 每秒取4帧画面左右
if cnt % int(fps / 4) == 0:
# 更新当前工程的检测进度
if pre_state is None:
state[0] = float(cur_time/ end)
else:
state[0] = min(0.9999, float(cur_time / end))
mainWindow.projectContext.nd_process = state[0]
mainWindow.projectContext.last_time = cur_time
subTitle = detect_subtitle(frame)
if subTitle is not None:
subTitle = normalize(subTitle)
# 第一次找到字幕
if lastSubTitle is None and subTitle is not None:
start_time = cur_time
# 字幕消失
elif lastSubTitle is not None and subTitle is None:
end_time = cur_time
res.append([start_time, end_time, lastSubTitle])
if (len(res) == 1 and res[-1][0] - last_time >= 1) or (len(res) > 1 and res[-1][0] - res[-2][1]) >= 1:
print('--------------------------------------------------')
recommend_lens = int((res[-1][0] - last_time) * normal_speed) if len(res) == 1 else int(
(res[-1][0] - res[-2][1]) * normal_speed)
# write_to_sheet(book_path, sheet_name, ['', '', '', '插入旁白，推荐字数为%d' % recommend_lens])
add_to_list(mainWindow, "旁白", ['', '', '', '插入旁白，推荐字数为%d' % recommend_lens])
print(start_time, end_time, lastSubTitle)
# write_to_sheet(book_path, sheet_name, [round(start_time, 2), round(end_time, 2), lastSubTitle, ''])
add_to_list(mainWindow, "字幕", [round(start_time, 3), round(end_time, 3), lastSubTitle, ''])
elif lastSubTitle is not None and subTitle is not None:
# 两句话连在一起，但是两句话不一样
if string_similar(lastSubTitle, subTitle) < 0.7:
end_time = cur_time
res.append([start_time, end_time, lastSubTitle])
if (len(res) == 1 and res[-1][0] - last_time >= 1) or (len(res) > 1 and res[-1][0] - res[-2][1]) >= 1:
print('--------------------------------------------------')
recommend_lens = int((res[-1][0] - last_time) * normal_speed) if len(res) == 1 else int(
(res[-1][0] - res[-2][1]) * normal_speed)
# write_to_sheet(book_path, sheet_name, ['', '', '', '插入旁白，推荐字数为%d' % recommend_lens])
add_to_list(mainWindow, "旁白", ['', '', '', '插入旁白，推荐字数为%d' % recommend_lens])
print(start_time, end_time, lastSubTitle)
# write_to_sheet(book_path, sheet_name, [round(start_time, 2), round(end_time, 2), lastSubTitle, ''])
add_to_list(mainWindow, "字幕", [round(start_time, 3), round(end_time, 3), lastSubTitle, ''])
start_time = end_time
else:
lastSubTitle = subTitle if len(subTitle) > len(lastSubTitle) else lastSubTitle
continue
# 当前字幕与上一段字幕不一样
lastSubTitle = subTitle
def add_to_list(mainWindow: MainWindow, element_type: str, li: list):
st_time_sec, ed_time_sec, subtitle, suggest = li
st_time_sec, ed_time_sec = str(st_time_sec), str(ed_time_sec)
# 默认使用配置文件中的语速
speed = mainWindow.projectContext.speaker_speed
aside = ""
i = len(mainWindow.projectContext.all_elements)
if element_type == "字幕":
mainWindow.projectContext.subtitle_list.append(Element(st_time_sec, ed_time_sec, subtitle, suggest, aside,speed))
mainWindow.projectContext.all_elements.append(mainWindow.projectContext.subtitle_list[-1])
else:
if i == 0:
st_time_sec = "0.01"
else:
st_time_sec = "%.2f" % (float(mainWindow.projectContext.all_elements[i - 1].ed_time_sec) + 0.01)
# 因为暂时没有用到ed_time_sec，所以直接赋值空吧
ed_time_sec = ""
mainWindow.projectContext.aside_list.append(Element(st_time_sec, ed_time_sec, subtitle, suggest, aside,speed))
mainWindow.projectContext.all_elements.append(mainWindow.projectContext.aside_list[-1])
def detect_with_ocr(video_path: str, book_path: str, start_time: float, end_time: float, state=None, mainWindow: MainWindow=None):
"""使用ocr检测视频获取字幕并输出旁白推荐
Args:
video_path (str): 待处理视频地址
book_path (str): 表格存储位置
start_time (float): 视频实际开始时间
end_time (float): 视频实际结束时间
state (optional): 用于通信的状态关键字. Defaults to None.
"""
print("start and end time:", start_time, end_time)
if state is None:
state = [None]
# if os.path.exists(book_path):
# os.remove(book_path)
book_name_xlsx = book_path
sheet_name_xlsx = "旁白插入位置建议"
context = mainWindow.projectContext
# 获取字幕在画面中的上下边界，方便在后续视频遍历过程中直接对字幕对应区域进行分析
global up_b, down_b
if context.detected:
up_b, down_b = context.caption_boundings[0], context.caption_boundings[1]
else:
# 此处start_time + 300是为了节省用户调整视频开始时间的功夫（强行跳过前5分钟）
up_b, down_b = get_position(video_path, start_time +300)
context.caption_boundings = [up_b, down_b]
context.detected = True
# 获取并构建输出信息
# table_head = [["起始时间", "终止时间", "字幕", '建议', '解说脚本']]
# print("create sheet at time: ", datetime.datetime.now())
# create_sheet(book_name_xlsx, sheet_name_xlsx, table_head)
# print("process the total video at time: ", datetime.datetime.now())
process_video(video_path, start_time, end_time, book_name_xlsx, sheet_name_xlsx, state, mainWindow)
if __name__ == '__main__':
path = "D:/mystudy/Eagle/accessibility_movie_1/test.mp4"
print("get_pos:", get_position(path, 0))
"""基于OCR的旁白区间检测算法
总体流程如下：
- 首先检测得到字幕的上下边界位置，框定大概的字幕检测范围；
- 视频每秒取3帧，并使用OCR获取字幕以及时间戳等信息；
- 根据字幕间的时间差判断能否插入旁白以及旁白的推荐字数。
外部调用方式：
.. code-block:: python
from detect_with_ocr import detect_with_ocr
detect_with_ocr(video_path, book_path, start_time, end_time, state)
"""
import os
import cv2
import numpy as np
from paddleocr import PaddleOCR
import sys
print("PaddleOCR load path:", os.path.abspath(sys.modules[PaddleOCR.__module__].__file__))
import difflib
import re
from typing import Tuple, Union
from detect_with_asr import create_sheet, write_to_sheet
from main_window import MainWindow, Element
# 字幕的上下边界
up_b, down_b = 0, 0
# 初始化ocr工具
ocr = PaddleOCR(use_angle_cls=True, lang="ch", show_log=False, use_gpu=False)
# 正常语速为4字/秒
normal_speed = 4
def get_position(video_path: str, start_time: float) -> Tuple[float, float]:
# return (885.0, 989.0)
"""根据对视频中的画面进行分析，确定字幕的位置，以便后续的字幕识别
Args:
video_path (str): 视频存储路径
start_time (float): 视频开始时间
Returns:
Tuple[float, float]: 字幕在整个画面中的上下边界位置
"""
print("video_path:", video_path)
video = cv2.VideoCapture(video_path)
# print("video:", video)
subtitle_position = {}
fps = video.get(cv2.CAP_PROP_FPS)
start = int(start_time * fps)
cnt = 0
txt_cnt = 0
pre_txt = None
video.set(cv2.CAP_PROP_POS_FRAMES, start)
height = int(video.get(cv2.CAP_PROP_FRAME_HEIGHT) * 0.6)
while True:
_, img = video.read()
# print("img:", img)
# gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
# cv2.imshow('img', gray)
# cv2.imshow(img)
cnt += 1
if img is None or cnt > 10000:
break
if cnt % int(fps / 3) != 0:
continue
img = img[height:]
res = ocr.ocr(img, cls=True)
sorted(res, key=lambda text: text[0][0][1])
bottom_position = None
if len(res) == 0:
continue
log = []
print("cnt:", cnt, "rect_num:", len(res))
for x in res:
# print("x:", x)
rect, (txt, confidence) = x
# font_size = rect[2][1] - rect[0][1]
mid = (rect[0][0] + rect[1][0]) / 2
gradient = np.arctan(abs((rect[1][1] - rect[0][1]) / (rect[1][0] - rect[0][0])))
# 可能是字幕的文本
if confidence > 0.9 and 0.4 * img.shape[1] < mid < 0.6 * img.shape[1] and gradient < 0.1:
if bottom_position is None:
bottom_position = rect[0][1]
# 判断是否与前一文本相同（是不是同一个字幕），非同一字幕的前提下，取对应上下边界，
keys = subtitle_position.keys()
if abs(rect[0][1] - bottom_position) < 10:
if pre_txt is None or pre_txt != txt:
txt_cnt += 1
pre_txt = txt
if (rect[0][0], rect[2][1]) in keys:
subtitle_position[(rect[0][1], rect[2][1])] += 1
else:
replace = False
for k in keys:
# 更新键值为最宽的上下限
if abs(rect[0][1] - k[0]) + abs(rect[2][1] - k[1]) < 10:
new_k = min(k[0], rect[0][1]), max(k[1], rect[2][1])
if new_k != k:
subtitle_position[new_k] = subtitle_position[k]
subtitle_position[new_k] += 1
subtitle_position.pop(k)
else:
subtitle_position[k] += 1
replace = True
break
if not replace:
subtitle_position[(rect[0][1], rect[2][1])] = 1
if txt_cnt == 3:
break
print(subtitle_position)
up_bounding, down_bounding = max(subtitle_position, key=subtitle_position.get)
return up_bounding + height, down_bounding + height
def erasePunc(txt: str) -> str:
"""去除字符串中的非中文字符
Args:
txt (str): 待处理字符串
Returns:
str: 处理后的字符串
"""
pattern = re.compile(r'[^\u4e00-\u9fa5]')
txt = re.sub(pattern, '', txt)
return txt
def string_similar(s1: str, s2: str) -> float:
"""比较字符串s1和s2的相似度，主要用于减少输出文件中相似字幕的重复
Args:
s1 (str): 第一个字符串
s2 (str): 第二个字符串
Returns:
float: 字符串间的相似度
"""
# 去除非中文字符后，再比较相似度
s1 = erasePunc(s1)
s2 = erasePunc(s2)
return difflib.SequenceMatcher(None, s1, s2).quick_ratio()
def normalize(text: str) -> str:
"""规范化处理文本中的一些标点符号
Args:
text (str): 待处理字符串
Returns:
str: 处理后的字符串
"""
# 将英文标点转换为中文标点
E_pun = u',.!?()[]:;'
C_pun = u'，。！？（）【】：；'
table = {ord(f): ord(t) for f, t in zip(E_pun, C_pun)}
text = text.translate(table)
text = text.strip(' ，。、【】_·：-@‘［；')
# 促成首尾匹配的（）
if text[-1] == '）' and text[0] != '（':
text = '（' + text
elif text[-1] != '）' and text[0] == '（':
text = text + '）'
return text
def detect_subtitle(img: np.ndarray) -> Union[str, None]:
"""检测当前画面得到字幕信息
Args:
img (np.ndarray): 当前画面
Returns:
Union[str, None]: 字幕信息(没有字幕时返回None)
"""
subTitle = ''
height = down_b - up_b
img = img[int(up_b - height * 0.7):int(down_b + height * 0.7)]
# 针对低帧率的视频做图像放大处理
if img.shape[1] < 1000:
img = cv2.resize(img, (int(img.shape[1] * 1.5), int(img.shape[0] * 1.5)))
res = ocr.ocr(img, cls=True)
sorted(res, key=lambda text: text[0][0][1])
if len(res) == 0:
return None
possible_txt = []
for x in res:
# cv2.imshow("cut", img)
# cv2.waitKey(0)
# cv2.destroyAllWindows()
rect, (txt, confidence) = x
font_size = rect[2][1] - rect[0][1]
mid = (rect[0][0] + rect[1][0]) / 2
gradient = np.arctan(abs((rect[1][1] - rect[0][1]) / (rect[1][0] - rect[0][0])))
# log.append("文本：{}，置信度：{}，中心点：{}，斜率：{}，字体大小：{}".format(txt, confidence, mid / img.shape[1], gradient,
# font_size)) 置信度>0.7 & 斜率<0.1 & 字幕偏移量<=25 & 字幕中心在画面宽的0.4-0.6之间
if confidence > 0.7 and gradient < 0.1 and 0.4 < mid / img.shape[1] < 0.6 and \
abs(rect[0][1] - 30) + abs(img.shape[0] - rect[2][1] - 30) <= 25:
subTitle += txt
# 如果字幕在一行中分为两个（或以上）对话文本
elif confidence > 0.85 and gradient < 0.1:
if 0.3 < mid / img.shape[1] < 0.4 or 0.6 < mid / img.shape[1] < 0.7:
possible_txt.append(txt)
if len(possible_txt) >= 2:
subTitle = ''.join(possible_txt)
if len(subTitle) > 0:
return subTitle
return None
def process_video(video_path: str, begin: float, end: float, book_path: str, sheet_name: str, state=None, mainWindow: MainWindow=None):
"""处理视频，主要完成对字幕的捕捉以及根据字幕分析得出旁白可能位置的任务
Args:
video_path (str): 待处理视频的路径
begin (float): 电影的实际开始位置（秒）
end (float): 电影除演职表外的实际结束位置（秒）
book_path (str): 输出表格地址
sheet_name (str): 输出表格中的表名
state (optional): 用于通信的状态关键字. Defaults to None.
"""
global normal_speed
if mainWindow.projectContext.speaker_speed is not None:
normal_speed = float(mainWindow.projectContext.speaker_speed.split('(')[1].split('字')[0])
sz = len(mainWindow.projectContext.all_elements)
if sz == 0:
last_time = begin
else:
last_time = float(mainWindow.projectContext.all_elements[sz - 1].ed_time_sec) + 0.01
print("当前使用的语速为", normal_speed)
if state is None:
state = [None]
video = cv2.VideoCapture(video_path)
fps = video.get(cv2.CAP_PROP_FPS)
lastSubTitle = None
# res是在视频遍历过程中获取的字幕文件，不掺杂对旁白的分析
res = []
cnt = 0
start_time = 0
end_time = 0
video.set(cv2.CAP_PROP_POS_MSEC, begin * 1000)
pre_state = state[0]
while True:
_, frame = video.read()
if frame is None:
break
cnt += 1
cur_time = video.get(cv2.CAP_PROP_POS_MSEC) / 1000
# 判断当前帧是否已超限制
if cur_time > end:
if cur_time - end_time > 1:
print('--------------------------------------------------')
recommend_lens = int((cur_time - end_time) * normal_speed)
# write_to_sheet(book_path, sheet_name, ['', '', '', '插入旁白，推荐字数为%d' % recommend_lens])
add_to_list(mainWindow, "旁白", ['', '', '', '插入旁白，推荐字数为%d' % recommend_lens])
# 判断当前是否有字幕需要被保存下来
if end_time < start_time:
# write_to_sheet(book_path, sheet_name, [round(start_time, 2), round(end, 2), lastSubTitle, ''])
add_to_list(mainWindow, "字幕", [round(start_time, 3), round(end, 3), lastSubTitle, ''])
break
# 每秒取4帧画面左右
if cnt % int(fps / 4) == 0:
# 更新当前工程的检测进度
if pre_state is None:
state[0] = float(cur_time/ end)
else:
state[0] = min(0.9999, float(cur_time / end))
mainWindow.projectContext.nd_process = state[0]
mainWindow.projectContext.last_time = cur_time
subTitle = detect_subtitle(frame)
if subTitle is not None:
subTitle = normalize(subTitle)
# 第一次找到字幕
if lastSubTitle is None and subTitle is not None:
start_time = cur_time
# 字幕消失
elif lastSubTitle is not None and subTitle is None:
end_time = cur_time
res.append([start_time, end_time, lastSubTitle])
if (len(res) == 1 and res[-1][0] - last_time >= 1) or (len(res) > 1 and res[-1][0] - res[-2][1]) >= 1:
print('--------------------------------------------------')
recommend_lens = int((res[-1][0] - last_time) * normal_speed) if len(res) == 1 else int(
(res[-1][0] - res[-2][1]) * normal_speed)
# write_to_sheet(book_path, sheet_name, ['', '', '', '插入旁白，推荐字数为%d' % recommend_lens])
add_to_list(mainWindow, "旁白", ['', '', '', '插入旁白，推荐字数为%d' % recommend_lens])
print(start_time, end_time, lastSubTitle)
# write_to_sheet(book_path, sheet_name, [round(start_time, 2), round(end_time, 2), lastSubTitle, ''])
add_to_list(mainWindow, "字幕", [round(start_time, 3), round(end_time, 3), lastSubTitle, ''])
elif lastSubTitle is not None and subTitle is not None:
# 两句话连在一起，但是两句话不一样
if string_similar(lastSubTitle, subTitle) < 0.7:
end_time = cur_time
res.append([start_time, end_time, lastSubTitle])
if (len(res) == 1 and res[-1][0] - last_time >= 1) or (len(res) > 1 and res[-1][0] - res[-2][1]) >= 1:
print('--------------------------------------------------')
recommend_lens = int((res[-1][0] - last_time) * normal_speed) if len(res) == 1 else int(
(res[-1][0] - res[-2][1]) * normal_speed)
# write_to_sheet(book_path, sheet_name, ['', '', '', '插入旁白，推荐字数为%d' % recommend_lens])
add_to_list(mainWindow, "旁白", ['', '', '', '插入旁白，推荐字数为%d' % recommend_lens])
print(start_time, end_time, lastSubTitle)
# write_to_sheet(book_path, sheet_name, [round(start_time, 2), round(end_time, 2), lastSubTitle, ''])
add_to_list(mainWindow, "字幕", [round(start_time, 3), round(end_time, 3), lastSubTitle, ''])
start_time = end_time
else:
lastSubTitle = subTitle if len(subTitle) > len(lastSubTitle) else lastSubTitle
continue
# 当前字幕与上一段字幕不一样
lastSubTitle = subTitle
def add_to_list(mainWindow: MainWindow, element_type: str, li: list):
st_time_sec, ed_time_sec, subtitle, suggest = li
st_time_sec, ed_time_sec = str(st_time_sec), str(ed_time_sec)
# 默认使用配置文件中的语速
speed = mainWindow.projectContext.speaker_speed
aside = ""
i = len(mainWindow.projectContext.all_elements)
if element_type == "字幕":
mainWindow.projectContext.subtitle_list.append(Element(st_time_sec, ed_time_sec, subtitle, suggest, aside,speed))
mainWindow.projectContext.all_elements.append(mainWindow.projectContext.subtitle_list[-1])
else:
if i == 0:
st_time_sec = "0.01"
else:
st_time_sec = "%.2f" % (float(mainWindow.projectContext.all_elements[i - 1].ed_time_sec) + 0.01)
# 因为暂时没有用到ed_time_sec，所以直接赋值空吧
ed_time_sec = ""
mainWindow.projectContext.aside_list.append(Element(st_time_sec, ed_time_sec, subtitle, suggest, aside,speed))
mainWindow.projectContext.all_elements.append(mainWindow.projectContext.aside_list[-1])
def detect_with_ocr(video_path: str, book_path: str, start_time: float, end_time: float, state=None, mainWindow: MainWindow=None):
"""使用ocr检测视频获取字幕并输出旁白推荐
Args:
video_path (str): 待处理视频地址
book_path (str): 表格存储位置
start_time (float): 视频实际开始时间
end_time (float): 视频实际结束时间
state (optional): 用于通信的状态关键字. Defaults to None.
"""
print("start and end time:", start_time, end_time)
if state is None:
state = [None]
# if os.path.exists(book_path):
# os.remove(book_path)
book_name_xlsx = book_path
sheet_name_xlsx = "旁白插入位置建议"
context = mainWindow.projectContext
# 获取字幕在画面中的上下边界，方便在后续视频遍历过程中直接对字幕对应区域进行分析
global up_b, down_b
if context.detected:
up_b, down_b = context.caption_boundings[0], context.caption_boundings[1]
else:
# 此处start_time + 300是为了节省用户调整视频开始时间的功夫（强行跳过前5分钟）
up_b, down_b = get_position(video_path, start_time +300)
context.caption_boundings = [up_b, down_b]
context.detected = True
# 获取并构建输出信息
# table_head = [["起始时间", "终止时间", "字幕", '建议', '解说脚本']]
# print("create sheet at time: ", datetime.datetime.now())
# create_sheet(book_name_xlsx, sheet_name_xlsx, table_head)
# print("process the total video at time: ", datetime.datetime.now())
process_video(video_path, start_time, end_time, book_name_xlsx, sheet_name_xlsx, state, mainWindow)
if __name__ == '__main__':
path = "D:/mystudy/Eagle/accessibility_movie_1/test.mp4"
print("get_pos:", get_position(path, 0))
main_window.py
View file @ 8142abd4
This source diff could not be displayed because it is too large. You can view the blob instead.
main_window_ui.py
View file @ 8142abd4
......@@ -21,6 +21,7 @@ class Ui_MainWindow(object):
" padding: 5px;\n"
"}")
self.centralwidget = QtWidgets.QWidget(MainWindow)
# self.centralwidget.setStyleSheet("font: 9pt \"黑体\";")
self.centralwidget.setObjectName("centralwidget")
self.horizontalLayout = QtWidgets.QHBoxLayout(self.centralwidget)
self.horizontalLayout.setObjectName("horizontalLayout")
......@@ -206,6 +207,7 @@ class Ui_MainWindow(object):
self.horizontalLayout_4 = QtWidgets.QHBoxLayout(self.all_tab)
self.horizontalLayout_4.setObjectName("horizontalLayout_4")
self.all_tableWidget = QtWidgets.QTableWidget(self.all_tab)
# self.all_tableWidget.setStyleSheet("font: 9pt \"微软雅黑\";")
self.all_tableWidget.setObjectName("all_tableWidget")
self.all_tableWidget.setColumnCount(0)
self.all_tableWidget.setRowCount(0)
......@@ -240,6 +242,7 @@ class Ui_MainWindow(object):
self.shuiping.setStretch(1, 5)
self.verticalLayout.addLayout(self.shuiping)
self.verticalWidget_2 = QtWidgets.QWidget(self.centralwidget)
# self.verticalWidget_2.setStyleSheet("font: 9pt \"微软雅黑\";")
self.verticalWidget_2.setObjectName("verticalWidget_2")
self.chuizhi = QtWidgets.QVBoxLayout(self.verticalWidget_2)
self.chuizhi.setObjectName("chuizhi")
......@@ -330,18 +333,26 @@ class Ui_MainWindow(object):
self.statusbar.setObjectName("statusbar")
MainWindow.setStatusBar(self.statusbar)
self.setting = QtWidgets.QAction(MainWindow)
# font = QtGui.QFont()
# font.setFamily("微软雅黑")
# self.setting.setFont(font)
self.setting.setObjectName("setting")
self.action_open_project = QtWidgets.QAction(MainWindow)
# self.action_open_project.setFont(font)
self.action_open_project.setObjectName("action_open_project")
self.import_movie = QtWidgets.QAction(MainWindow)
# self.import_movie.setFont(font)
self.import_movie.setObjectName("import_movie")
self.actions = QtWidgets.QAction(MainWindow)
self.actions.setObjectName("actions")
self.action_save = QtWidgets.QAction(MainWindow)
# self.action_save.setFont(font)
self.action_save.setObjectName("action_save")
self.action_undo = QtWidgets.QAction(MainWindow)
# self.action_undo.setFont(font)
self.action_undo.setObjectName("action_undo")
self.action_redo = QtWidgets.QAction(MainWindow)
# self.action_redo.setFont(font)
self.action_redo.setObjectName("action_redo")
self.action_3 = QtWidgets.QAction(MainWindow)
self.action_3.setObjectName("action_3")
......@@ -352,10 +363,12 @@ class Ui_MainWindow(object):
self.action_operate = QtWidgets.QAction(MainWindow)
self.action_operate.setObjectName("action_operate")
self.action_export = QtWidgets.QAction(MainWindow)
# self.action_export.setFont(font)
self.action_export.setObjectName("action_export")
self.action_insert_aside_from_now = QtWidgets.QAction(MainWindow)
self.action_insert_aside_from_now.setObjectName("action_insert_aside_from_now")
self.action_create = QtWidgets.QAction(MainWindow)
# self.action_create.setFont(font)
self.action_create.setObjectName("action_create")
self.menu.addAction(self.action_create)
self.menu.addAction(self.action_open_project)
......
management.py
View file @ 8142abd4
......@@ -174,6 +174,7 @@ class ProjectContext:
return
with open(self.conf_path, 'r', encoding='utf8') as f:
info = json.load(f)
# print(json.dumps(info, ensure_ascii=False, indent=4))
self.video_path = info["video_path"]
self.excel_path = info["excel_path"]
self.speaker_info = info["speaker_info"]["speaker_id"]
......@@ -207,8 +208,7 @@ class ProjectContext:
}
}
f.write(json.dumps(info))
def setVideoPath(self, video_path):
self.video_path = video_path
......
operation_dialog.py
View file @ 8142abd4
......@@ -141,7 +141,7 @@ class Operation_Dialog(QDialog, Ui_Dialog):
assert start_time_f < end_time_f
if row_number < rowCount:
# 要求结束时间在下一行的起始时间前面，对增加和修改都适用
assert float(self.mainWindow.projectContext.all_elements[list_idx+1].st_time_sec) > end_time_f
assert float(self.mainWindow.projectContext.all_elements[list_idx+1].st_time_sec) >= end_time_f
# 要求起始时间在上一行后面，在下一行前面
if self.comboBox_2.currentText() == "增加一行":
......
operation_dialog_ui.py
View file @ 8142abd4
# -*- coding: utf-8 -*-
# Form implementation generated from reading ui file 'operation_dialog.ui'
# Form implementation generated from reading ui file 'd:\AddCaption\cur_version\accessibility_movie_2\operation_dialog.ui'
#
# Created by: PyQt5 UI code generator 5.12
# Created by: PyQt5 UI code generator 5.15.4
#
# WARNING! All changes made in this file will be lost!
# WARNING: Any manual changes made to this file will be lost when pyuic5 is
# run again. Do not edit this file unless you know what you are doing.
from PyQt5 import QtCore, QtGui, QtWidgets
......@@ -12,52 +14,20 @@ from PyQt5 import QtCore, QtGui, QtWidgets
class Ui_Dialog(object):
def setupUi(self, Dialog):
Dialog.setObjectName("Dialog")
Dialog.resize(724, 437)
Dialog.resize(724, 502)
self.gridLayout = QtWidgets.QGridLayout(Dialog)
self.gridLayout.setObjectName("gridLayout")
self.label_14 = QtWidgets.QLabel(Dialog)
self.label_14.setObjectName("label_14")
self.gridLayout.addWidget(self.label_14, 5, 3, 1, 1)
self.label_11 = QtWidgets.QLabel(Dialog)
self.label_11.setObjectName("label_11")
self.gridLayout.addWidget(self.label_11, 6, 2, 1, 1)
self.comboBox_3 = QtWidgets.QComboBox(Dialog)
self.comboBox_3.setEnabled(True)
self.comboBox_3.setEditable(False)
self.comboBox_3.setObjectName("comboBox_3")
self.gridLayout.addWidget(self.comboBox_3, 5, 4, 1, 1)
self.label_11 = QtWidgets.QLabel(Dialog)
self.label_11.setObjectName("label_11")
self.gridLayout.addWidget(self.label_11, 5, 2, 1, 1)
self.gridLayout.addWidget(self.comboBox_3, 6, 4, 1, 1)
self.label_9 = QtWidgets.QLabel(Dialog)
self.label_9.setObjectName("label_9")
self.gridLayout.addWidget(self.label_9, 3, 2, 1, 1)
self.lineEdit_6 = QtWidgets.QLineEdit(Dialog)
self.lineEdit_6.setEnabled(True)
self.lineEdit_6.setObjectName("lineEdit_6")
self.gridLayout.addWidget(self.lineEdit_6, 5, 1, 1, 1)
self.lineEdit_4 = QtWidgets.QLineEdit(Dialog)
self.lineEdit_4.setEnabled(True)
self.lineEdit_4.setObjectName("lineEdit_4")
self.gridLayout.addWidget(self.lineEdit_4, 4, 1, 1, 1)
self.pickEndPos = QtWidgets.QPushButton(Dialog)
self.pickEndPos.setObjectName("pickEndPos")
self.gridLayout.addWidget(self.pickEndPos, 3, 3, 1, 1)
self.pickStartPos = QtWidgets.QPushButton(Dialog)
self.pickStartPos.setObjectName("pickStartPos")
self.gridLayout.addWidget(self.pickStartPos, 2, 3, 1, 1)
self.label_4 = QtWidgets.QLabel(Dialog)
self.label_4.setObjectName("label_4")
self.gridLayout.addWidget(self.label_4, 3, 0, 1, 1)
self.label = QtWidgets.QLabel(Dialog)
self.label.setObjectName("label")
self.gridLayout.addWidget(self.label, 1, 0, 1, 1)
self.pushButton_3 = QtWidgets.QPushButton(Dialog)
self.pushButton_3.setObjectName("pushButton_3")
self.gridLayout.addWidget(self.pushButton_3, 1, 3, 1, 2)
self.label_3 = QtWidgets.QLabel(Dialog)
self.label_3.setObjectName("label_3")
self.gridLayout.addWidget(self.label_3, 2, 0, 1, 1)
self.label_8 = QtWidgets.QLabel(Dialog)
self.label_8.setObjectName("label_8")
self.gridLayout.addWidget(self.label_8, 2, 2, 1, 1)
self.verticalLayout = QtWidgets.QVBoxLayout()
self.verticalLayout.setObjectName("verticalLayout")
self.label_2 = QtWidgets.QLabel(Dialog)
......@@ -70,9 +40,41 @@ class Ui_Dialog(object):
self.label_13.setObjectName("label_13")
self.verticalLayout.addWidget(self.label_13)
self.gridLayout.addLayout(self.verticalLayout, 1, 2, 1, 1)
self.label_4 = QtWidgets.QLabel(Dialog)
self.label_4.setObjectName("label_4")
self.gridLayout.addWidget(self.label_4, 3, 0, 1, 1)
self.pickStartPos = QtWidgets.QPushButton(Dialog)
self.pickStartPos.setObjectName("pickStartPos")
self.gridLayout.addWidget(self.pickStartPos, 2, 3, 1, 1)
self.label_10 = QtWidgets.QLabel(Dialog)
self.label_10.setObjectName("label_10")
self.gridLayout.addWidget(self.label_10, 4, 2, 1, 1)
self.label_14 = QtWidgets.QLabel(Dialog)
self.label_14.setObjectName("label_14")
self.gridLayout.addWidget(self.label_14, 6, 3, 1, 1)
self.label_7 = QtWidgets.QLabel(Dialog)
self.label_7.setObjectName("label_7")
self.gridLayout.addWidget(self.label_7, 5, 0, 1, 1)
self.gridLayout.addWidget(self.label_7, 6, 0, 1, 1)
self.label_3 = QtWidgets.QLabel(Dialog)
self.label_3.setObjectName("label_3")
self.gridLayout.addWidget(self.label_3, 2, 0, 1, 1)
self.pickEndPos = QtWidgets.QPushButton(Dialog)
self.pickEndPos.setObjectName("pickEndPos")
self.gridLayout.addWidget(self.pickEndPos, 3, 3, 1, 1)
self.lineEdit_6 = QtWidgets.QLineEdit(Dialog)
self.lineEdit_6.setEnabled(True)
self.lineEdit_6.setObjectName("lineEdit_6")
self.gridLayout.addWidget(self.lineEdit_6, 6, 1, 1, 1)
self.lineEdit_4 = QtWidgets.QLineEdit(Dialog)
self.lineEdit_4.setEnabled(True)
self.lineEdit_4.setObjectName("lineEdit_4")
self.gridLayout.addWidget(self.lineEdit_4, 4, 1, 1, 1)
self.label = QtWidgets.QLabel(Dialog)
self.label.setObjectName("label")
self.gridLayout.addWidget(self.label, 1, 0, 1, 1)
self.pushButton_3 = QtWidgets.QPushButton(Dialog)
self.pushButton_3.setObjectName("pushButton_3")
self.gridLayout.addWidget(self.pushButton_3, 1, 3, 1, 2)
self.buttonBox = QtWidgets.QDialogButtonBox(Dialog)
self.buttonBox.setEnabled(False)
sizePolicy = QtWidgets.QSizePolicy(QtWidgets.QSizePolicy.Minimum, QtWidgets.QSizePolicy.Fixed)
......@@ -87,10 +89,31 @@ class Ui_Dialog(object):
self.buttonBox.setStandardButtons(QtWidgets.QDialogButtonBox.Ok)
self.buttonBox.setCenterButtons(False)
self.buttonBox.setObjectName("buttonBox")
self.gridLayout.addWidget(self.buttonBox, 7, 4, 1, 1)
self.label_10 = QtWidgets.QLabel(Dialog)
self.label_10.setObjectName("label_10")
self.gridLayout.addWidget(self.label_10, 4, 2, 1, 1)
self.gridLayout.addWidget(self.buttonBox, 8, 4, 1, 1)
self.label_8 = QtWidgets.QLabel(Dialog)
self.label_8.setObjectName("label_8")
self.gridLayout.addWidget(self.label_8, 2, 2, 1, 1)
self.horizontalLayout_2 = QtWidgets.QHBoxLayout()
self.horizontalLayout_2.setContentsMargins(-1, 0, -1, -1)
self.horizontalLayout_2.setObjectName("horizontalLayout_2")
self.pushButton_2 = QtWidgets.QPushButton(Dialog)
self.pushButton_2.setObjectName("pushButton_2")
self.horizontalLayout_2.addWidget(self.pushButton_2)
spacerItem = QtWidgets.QSpacerItem(40, 20, QtWidgets.QSizePolicy.Expanding, QtWidgets.QSizePolicy.Minimum)
self.horizontalLayout_2.addItem(spacerItem)
self.pushButton = QtWidgets.QPushButton(Dialog)
self.pushButton.setObjectName("pushButton")
self.horizontalLayout_2.addWidget(self.pushButton)
self.horizontalLayout_2.setStretch(0, 1)
self.horizontalLayout_2.setStretch(1, 1)
self.horizontalLayout_2.setStretch(2, 1)
self.gridLayout.addLayout(self.horizontalLayout_2, 8, 1, 1, 2)
self.label_6 = QtWidgets.QLabel(Dialog)
self.label_6.setObjectName("label_6")
self.gridLayout.addWidget(self.label_6, 5, 0, 1, 1)
self.label_5 = QtWidgets.QLabel(Dialog)
self.label_5.setObjectName("label_5")
self.gridLayout.addWidget(self.label_5, 4, 0, 1, 1)
self.horizontalLayout = QtWidgets.QHBoxLayout()
self.horizontalLayout.setObjectName("horizontalLayout")
self.comboBox = QtWidgets.QComboBox(Dialog)
......@@ -100,42 +123,29 @@ class Ui_Dialog(object):
self.comboBox.addItem("")
self.comboBox.addItem("")
self.horizontalLayout.addWidget(self.comboBox)
spacerItem = QtWidgets.QSpacerItem(40, 20, QtWidgets.QSizePolicy.Expanding, QtWidgets.QSizePolicy.Minimum)
self.horizontalLayout.addItem(spacerItem)
spacerItem1 = QtWidgets.QSpacerItem(40, 20, QtWidgets.QSizePolicy.Expanding, QtWidgets.QSizePolicy.Minimum)
self.horizontalLayout.addItem(spacerItem1)
self.horizontalLayout.setStretch(0, 1)
self.horizontalLayout.setStretch(1, 1)
self.gridLayout.addLayout(self.horizontalLayout, 0, 1, 1, 1)
self.label_5 = QtWidgets.QLabel(Dialog)
self.label_5.setObjectName("label_5")
self.gridLayout.addWidget(self.label_5, 4, 0, 1, 1)
self.timeEdit = QtWidgets.QTimeEdit(Dialog)
self.timeEdit.setCurrentSection(QtWidgets.QDateTimeEdit.MinuteSection)
self.timeEdit.setCurrentSectionIndex(1)
self.timeEdit.setObjectName("timeEdit")
self.gridLayout.addWidget(self.timeEdit, 2, 1, 1, 1)
self.horizontalLayout_2 = QtWidgets.QHBoxLayout()
self.horizontalLayout_2.setContentsMargins(-1, 0, -1, -1)
self.horizontalLayout_2.setObjectName("horizontalLayout_2")
self.pushButton_2 = QtWidgets.QPushButton(Dialog)
self.pushButton_2.setObjectName("pushButton_2")
self.horizontalLayout_2.addWidget(self.pushButton_2)
spacerItem1 = QtWidgets.QSpacerItem(40, 20, QtWidgets.QSizePolicy.Expanding, QtWidgets.QSizePolicy.Minimum)
self.horizontalLayout_2.addItem(spacerItem1)
self.pushButton = QtWidgets.QPushButton(Dialog)
self.pushButton.setObjectName("pushButton")
self.horizontalLayout_2.addWidget(self.pushButton)
self.horizontalLayout_2.setStretch(0, 1)
self.horizontalLayout_2.setStretch(1, 1)
self.gridLayout.addLayout(self.horizontalLayout_2, 7, 1, 1, 2)
self.comboBox_2 = QtWidgets.QComboBox(Dialog)
self.comboBox_2.setObjectName("comboBox_2")
self.comboBox_2.addItem("")
self.comboBox_2.addItem("")
self.comboBox_2.addItem("")
self.gridLayout.addWidget(self.comboBox_2, 0, 2, 1, 1)
self.lineEdit_5 = QtWidgets.QLineEdit(Dialog)
self.lineEdit_5.setEnabled(True)
self.lineEdit_5.setObjectName("lineEdit_5")
self.gridLayout.addWidget(self.lineEdit_5, 5, 1, 1, 1)
self.lineEdit = QtWidgets.QLineEdit(Dialog)
self.lineEdit.setObjectName("lineEdit")
self.gridLayout.addWidget(self.lineEdit, 1, 1, 1, 1)
self.timeEdit = QtWidgets.QTimeEdit(Dialog)
self.timeEdit.setCurrentSection(QtWidgets.QDateTimeEdit.MinuteSection)
self.timeEdit.setCurrentSectionIndex(1)
self.timeEdit.setObjectName("timeEdit")
self.gridLayout.addWidget(self.timeEdit, 2, 1, 1, 1)
self.timeEdit_2 = QtWidgets.QTimeEdit(Dialog)
self.timeEdit_2.setCurrentSection(QtWidgets.QDateTimeEdit.SecondSection)
self.timeEdit_2.setCurrentSectionIndex(2)
......@@ -148,30 +158,29 @@ class Ui_Dialog(object):
def retranslateUi(self, Dialog):
_translate = QtCore.QCoreApplication.translate
Dialog.setWindowTitle(_translate("Dialog", "Dialog"))
self.label_14.setText(_translate("Dialog", "语速："))
self.label_11.setText(_translate("Dialog", "*请填文字"))
self.label_9.setText(_translate("Dialog", "*请填数字，最多保留两位小数"))
self.pickEndPos.setText(_translate("Dialog", "取当前位置"))
self.pickStartPos.setText(_translate("Dialog", "取当前位置"))
self.label_2.setText(_translate("Dialog", "行（如果操作是增加，则在该行后面增加）"))
self.label_13.setText(_translate("Dialog", "*需要填在【字幕旁白】页面中的行数"))
self.label_4.setText(_translate("Dialog", "结束时间："))
self.pickStartPos.setText(_translate("Dialog", "取当前位置"))
self.label_10.setText(_translate("Dialog", "*请填文字"))
self.label_14.setText(_translate("Dialog", "语速："))
self.label_7.setText(_translate("Dialog", "旁白："))
self.label_3.setText(_translate("Dialog", "起始时间："))
self.pickEndPos.setText(_translate("Dialog", "取当前位置"))
self.label.setText(_translate("Dialog", "我想操作第"))
self.pushButton_3.setText(_translate("Dialog", "填充\n"
"行信息"))
self.label_3.setText(_translate("Dialog", "起始时间："))
self.label_8.setText(_translate("Dialog", "*请填数字，最多保留两位小数"))
self.label_2.setText(_translate("Dialog", "行（如果操作是增加，则在该行后面增加）"))
self.label_13.setText(_translate("Dialog", "*需要填在【字幕旁白】页面中的行数"))
self.label_7.setText(_translate("Dialog", "旁白："))
self.label_10.setText(_translate("Dialog", "*请填文字"))
self.comboBox.setItemText(0, _translate("Dialog", "字幕"))
self.comboBox.setItemText(1, _translate("Dialog", "旁白"))
self.label_5.setText(_translate("Dialog", "字幕："))
self.timeEdit.setDisplayFormat(_translate("Dialog", "hh:mm:ss.zzz"))
self.pushButton_2.setText(_translate("Dialog", "修改"))
self.pushButton.setText(_translate("Dialog", "检测"))
self.label_6.setText(_translate("Dialog", "推荐字数："))
self.label_5.setText(_translate("Dialog", "字幕："))
self.comboBox.setItemText(0, _translate("Dialog", "字幕"))
self.comboBox.setItemText(1, _translate("Dialog", "旁白"))
self.comboBox_2.setItemText(0, _translate("Dialog", "增加一行"))
self.comboBox_2.setItemText(1, _translate("Dialog", "修改一行"))
self.comboBox_2.setItemText(2, _translate("Dialog", "删除一行"))
self.timeEdit.setDisplayFormat(_translate("Dialog", "hh:mm:ss.zzz"))
self.timeEdit_2.setDisplayFormat(_translate("Dialog", "hh:mm:ss.zzz"))
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment