Caffe2 - C++ API
A deep learning, cross platform ML framework
Public Member Functions | Data Fields
caffe2::Params Class Reference

Public Member Functions

Paramsfps (float v)
 FPS of output frames setting here will reset intervals_ and force decoding at target FPS This can be used if user just want to decode at a steady fps.
ParamssetSampleTimestamps (const std::vector< double > &timestamps)
 Sample output frames at a specified list of timestamps Timestamps must be in increasing order, and timestamps past the end of the video will be ignored Setting here will reset intervals_.
ParamspixelFormat (AVPixelFormat pixelFormat)
 Pixel format of output buffer, default PIX_FMT_RGB24.
ParamskeyFrames (bool keyFrames)
 Return all key-frames.
ParamsstreamIndex (int index)
 Index of video stream to process, defaults to the first video stream.
ParamsmaxOutputFrames (int count)
 Only output this many frames, default to no limit.
ParamsoutputWidth (int width)
 Output frame width, default to video width.
ParamsoutputHeight (int height)
 Output frame height, default to video height.

Data Fields

bool keyFrames_ = false
AVPixelFormat pixelFormat_ = AVPixelFormat::AV_PIX_FMT_RGB24
int streamIndex_ = -1
int maximumOutputFrames_ = -1
int video_res_type_ = VideoResType::USE_WIDTH_HEIGHT
int crop_height_ = -1
int crop_width_ = -1
int height_min_ = -1
int width_min_ = -1
int scale_w_ = -1
int scale_h_ = -1
int decode_type_ = DecodeType::DO_TMP_JITTER
int num_of_required_frame_ = -1
std::vector< SampleIntervalintervals_ = {{0, SpecialFps::SAMPLE_ALL_FRAMES}}

Detailed Description

Definition at line 70 of file video_decoder.h.

The documentation for this class was generated from the following file: