Ask Your Question

fommilie's profile - activity

2017-05-18 12:26:01 -0500 received badge  Supporter (source)
2017-05-18 12:23:24 -0500 received badge  Student (source)
2017-05-18 12:07:46 -0500 asked a question butteraugli implementation for image difference?

Will an implementation of the Google psychovisual difference between two images be included in OpenCV https://github.com/google/butteraugli ?

I'd use it directly but I'm accessing things via the Java API and an OpenCV binding would save me having to create one.

2017-05-18 12:04:13 -0500 received badge  Editor (source)
2017-05-18 12:03:13 -0500 asked a question translate example into Java

I'm trying to translate this demo into Java http://docs.opencv.org/3.2.0/d5/dc4/t...

but I'm stuck on the lines using overloading of -= on Mat. My understanding of C++ is very limited, but I expected to see an operator-=()here http://docs.opencv.org/3.2.0/d3/d63/c... but there isn't one. What is the Java equivalent of -= on a Mat?

Note that the Java docs for 3.2.0 are not available, so consult 3.1.0 instead http://docs.opencv.org/java/3.1.0/

2017-05-17 11:55:05 -0500 marked best answer VideoCapture: partially decoded video data?

I've been following the VideoCapture tutorial http://docs.opencv.org/3.2.0/d5/dc4/t... and it strikes me that only fully decoded frames are analysed.

Is there a way to get hold of lower level information, i.e. partial decodings of an mp4 file? Since this is all I need for my analysis, I hope it would be more efficient than decoding the entire video:

  1. key frames (potentially in their encoded form, but decoded is also fine)
  2. p-frame motion vectors... I know h264 gets a bit more complex with full inter-frame but just forward motion estimation is fine.
  3. the fourier space from AAC (or anything that would resemble the DTFT... I realise this is a bit more ill-defined, it's not as important)

thanks!