Recherche avancée

Sur d’autres sites (9724)

• ffplay : support rotated video.

  18 avril 2014, par Clément Bœsch

  ffplay : support rotated video.
  with -f lavfi -i testsrc=s=hd1080 as source :
  
    rotate=90*PI/180  vs transpose=clock :   42fps -> 64fps
  
    rotate=180*PI/180 vs vflip,hflip :       75fps -> 77fps
  
    rotate=270*PI/180 vs transpose=cclock :  43fps -> 63fps
  

  • [DH] configure
  • [DH] doc/ffplay.texi
  • [DH] ffplay.c

• FFMPEG Scene Detection - Trim Blank Background from Video at the Start and End

  10 juin 2014, par user3521682

  Summary :
  Need to programatically trim video when scene is not changing at the beginning & end.

  Example video : http://www.filehosting.co.nz/finished3.mp4
  (Quality is much higher in the real video)

  Background :

  Large number of videos for online store, each video begins with a blank background, then the model walks on (at a random time, few seconds), and then walks off after a random time (around 15 seconds). The end of the video is trimmed seemingly random ; could be up to 15 seconds of ’nothing’ at the end of the video.

  The camera does not move. There is no sound on the videos.
  The videos come from a camera in MOV format, sideways.

  I already have FFMPEG converting from MOV to MP4, rotating the video, adding an audio-track, and trimming the audio at the end of the video.

  Research :
  I understand that I should probably re-encode video with a very high (?) tolerance for i-frames (so that only two are made per video) and then export the times to a text file, and use the text file to cut the video (probably parse it in BASH and use that to build the FFMPEG commands).

  Does anyone have any idea how I could generate just two key-frames per video ?

  Example video : http://www.filehosting.co.nz/finished3.mp4
  (Quality is much higher in the real video)

• avfilter/dnn : Refactor DNN parameter configuration system

  7 mai 2024, par Zhao Zhili

  avfilter/dnn : Refactor DNN parameter configuration system
  This patch trying to resolve mulitiple issues related to parameter
  
  configuration :
  Firstly, each DNN filters duplicate DNN_COMMON_OPTIONS, which should
  
  be the common options of backend.
  Secondly, backend options are hidden behind the scene. It's a
  
  AV_OPT_TYPE_STRING backend_configs for user, and parsed by each
  
  backend. We don't know each backend support what kind of options
  
  from the help message.
  Third, DNN backends duplicate DNN_BACKEND_COMMON_OPTIONS.
  Last but not the least, pass backend options via AV_OPT_TYPE_STRING
  
  makes it hard to pass AV_OPT_TYPE_BINARY to backend, if not impossible.
  This patch puts backend common options and each backend options inside
  
  DnnContext to reduce code duplication, make options user friendly, and
  
  easy to extend for future usecase.
  For example,
  ./ffmpeg -h filter=dnn_processing
  dnn_processing AVOptions :
  
     dnn_backend       <int>        ..FV....... DNN backend (from INT_MIN to INT_MAX) (default tensorflow)
  
       tensorflow      1            ..FV....... tensorflow backend flag
  
       openvino        2            ..FV....... openvino backend flag
  
       torch           3            ..FV....... torch backend flag
  dnn_base AVOptions :
  
     model             <string>     ..F........ path to model file
  
     input             <string>     ..F........ input name of the model
  
     output            <string>     ..F........ output name of the model
  
     backend_configs   <string>     ..F.......P backend configs (deprecated)
  
     options           <string>     ..F.......P backend configs (deprecated)
  
     nireq             <int>        ..F........ number of request (from 0 to INT_MAX) (default 0)
  
     async             <boolean>    ..F........ use DNN async inference (default true)
  
     device            <string>     ..F........ device to run model
  dnn_tensorflow AVOptions :
  
     sess_config       <string>     ..F........ config for SessionOptions
  dnn_openvino AVOptions :
  
     batch_size        <int>        ..F........ batch size per request (from 1 to 1000) (default 1)
  
     input_resizable   <boolean>    ..F........ can input be resizable or not (default false)
  
     layout            <int>        ..F........ input layout of model (from 0 to 2) (default none)
  
       none            0            ..F........ none
  
       nchw            1            ..F........ nchw
  
       nhwc            2            ..F........ nhwc
  
     scale             <float>      ..F........ Add scale preprocess operation. Divide each element of input by specified value. (from INT_MIN to INT_MAX) (default 0)
  
     mean              <float>      ..F........ Add mean preprocess operation. Subtract specified value from each element of input. (from INT_MIN to INT_MAX) (default 0)
  dnn_th AVOptions :
  
     optimize          <int>        ..F........ turn on graph executor optimization (from 0 to 1) (default 0)
  Signed-off-by : Zhao Zhili <zhilizhao@tencent.com>
  
  Reviewed-by : Wenbin Chen <wenbin.chen@intel.com>
  
  Reviewed-by : Guo Yejun <yejun.guo@intel.com>
  

  • [DH] libavfilter/dnn/dnn_backend_common.h
  • [DH] libavfilter/dnn/dnn_backend_openvino.c
  • [DH] libavfilter/dnn/dnn_backend_tf.c
  • [DH] libavfilter/dnn/dnn_backend_torch.cpp
  • [DH] libavfilter/dnn/dnn_interface.c
  • [DH] libavfilter/dnn_filter_common.c
  • [DH] libavfilter/dnn_filter_common.h
  • [DH] libavfilter/dnn_interface.h
  • [DH] libavfilter/vf_derain.c
  • [DH] libavfilter/vf_dnn_classify.c
  • [DH] libavfilter/vf_dnn_detect.c
  • [DH] libavfilter/vf_dnn_processing.c
  • [DH] libavfilter/vf_sr.c