Update 3mergeDialogToVideo.py

3mergeDialogToVideo.py

@@ -273,6 +273,41 @@ if __name__ == '__main__':
 
     image_files = filtered_image_files
 
+    
+
+    # for idx, image_file in enumerate(image_files):
+    #     print("现在处理的图片是"+image_file)
+    #     #后面是视音频生成部分-这里图片需要用到完整的去水印的而不是对话框用于识别的
+    #     img = cv2.imread(image_file)
+    #     img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)  ##只支持英文路径
+
+    #     ##获取当前图片对应的对话框识别文字(还需gpt处理后作为字幕文案)
+    #     cur_copywrite = get_image_copywrite(image_file,dialog_img_path)  # image_file就是6.jpg了
+    #     #cur_copywrite = gpt_polish(cur_copywrite)#不用gpt，只用新版漫画块得到的100%识别原文即可
+
+    #     if cur_copywrite is not None:
+
+    #         ##获取当前图片对应的临时音频文件名称和文案时长
+    #         filename, duration = get_audio_data(cur_copywrite)
+    #         if filename is not None:
+    #             print("存放临时mp3文件的路径是",filename)
+    #             clip = ImageClip(img).set_duration(duration).resize((width, height))  # 初始clip
+    
+    #             txt_clip = TextClip(cur_copywrite, fontsize=40, color='white', bg_color='black',
+    #                                 font=font_path)  ##文本clip后加入视频
+    
+    #             txt_clip = txt_clip.set_pos(('center', 'bottom')).set_duration(duration)
+    #             # 创建音频剪辑
+    #             audio_clip = AudioFileClip(filename)
+    #             clip = clip.set_audio(audio_clip)  # 将音频与视频片段关联
+    #             clip = CompositeVideoClip([clip, txt_clip])
+    #             video_clips.append(clip)
+    #         else:
+    #             pass ##音频特殊字符或者其他原因无法生成跳过
+
+
+    #分割字幕
+    prev_clip_end = 0
     for idx, image_file in enumerate(image_files):
         print("现在处理的图片是"+image_file)
         #后面是视音频生成部分-这里图片需要用到完整的去水印的而不是对话框用于识别的
@@ -295,15 +330,21 @@ if __name__ == '__main__':
                                     font=font_path)  ##文本clip后加入视频
     
                 txt_clip = txt_clip.set_pos(('center', 'bottom')).set_duration(duration)
+                if prev_clip_end > 0:
+                    silence_duration = duration - prev_clip_end
+                    silence_clip = TextClip('', fontsize=40, color='white', bg_color='black', font=font_path)
+                    silence_clip = silence_clip.set_pos(('center', 'bottom')).set_duration(silence_duration)
+                    video_clips.append(silence_clip)
+                
                 # 创建音频剪辑
                 audio_clip = AudioFileClip(filename)
                 clip = clip.set_audio(audio_clip)  # 将音频与视频片段关联
                 clip = CompositeVideoClip([clip, txt_clip])
                 video_clips.append(clip)
+                prev_clip_end = duration
             else:
                 pass ##音频特殊字符或者其他原因无法生成跳过
-        else:
-            pass  ##图片不规范直接跳过
+        
     video = concatenate_videoclips(video_clips)
     # 保存视频
     video.write_videofile('mp4_out/output_video.mp4', fps=fps,temp_audiofile="mp3_out/temp.mp3")