Merge pull request #86 from zhangtiance/master

Support video without audio.

util/ffmpeg.py

@@ -52,7 +52,10 @@ def video2voice(videopath, voicepath, start_time='00:00:00', last_time='00:00:00
 
 def image2video(fps,imagepath,voicepath,videopath):
     os.system('ffmpeg -y -r '+str(fps)+' -i '+imagepath+' -vcodec libx264 '+os.path.split(voicepath)[0]+'/video_tmp.mp4')
-    os.system('ffmpeg -i '+os.path.split(voicepath)[0]+'/video_tmp.mp4'+' -i "'+voicepath+'" -vcodec copy -acodec aac '+videopath)
+    if os.path.exists(voicepath):
+        os.system('ffmpeg -i '+os.path.split(voicepath)[0]+'/video_tmp.mp4'+' -i "'+voicepath+'" -vcodec copy -acodec aac '+videopath)
+    else:
+        os.system('ffmpeg -i '+os.path.split(voicepath)[0]+'/video_tmp.mp4 '+videopath)
 
 def get_video_infos(videopath):
     args =  ['ffprobe -v quiet -print_format json -show_format -show_streams', '-i', '"'+videopath+'"']