Borrar filtros
Borrar filtros

Use RNN to detect key frames in a video

2 visualizaciones (últimos 30 días)
Hossam Elshahaby
Hossam Elshahaby el 24 de Sept. de 2018
Comentada: Walter Roberson el 24 de Sept. de 2018
Hello Mr/Ms,
I want to use RNN network to get sequence of frames and output another sequence of key frames.
How can I achieve that purpose?
Thanks,
  3 comentarios
Hossam Elshahaby
Hossam Elshahaby el 24 de Sept. de 2018
Hi Walter, My application is OCR for text in videos which is rotating horizontally or flipping vertically . I want to detect the frames with totally new text .
What are the optimum algorithm that I can use?
Walter Roberson
Walter Roberson el 24 de Sept. de 2018
That is an algorithm question, not a MATLAB question. It is also a different algorithm question than was originally asked as key frame detection does not have much to do with ocr.

Iniciar sesión para comentar.

Respuestas (1)

Walter Roberson
Walter Roberson el 24 de Sept. de 2018
MATLAB has an ocr() function. You can train the underlying neural network with text in a variety of orientations.
  2 comentarios
Hossam Elshahaby
Hossam Elshahaby el 24 de Sept. de 2018
Editada: Hossam Elshahaby el 24 de Sept. de 2018
Hello Walter,
I have sequence of frames for news caption images cropped from real news video
for example: 6 frames
  • Obama will visit Ger
  • will visit Germany
  • l visit Germany next
  • Visit Germany next week
  • ....text rotate....
  • He will sign contracts
The first and last one are unique (key frames)
This should be the output of the network.
How can I achieve that? What are the best algorithms to use?
Thanks
Walter Roberson
Walter Roberson el 24 de Sept. de 2018
This is resource for asking questions about how MATLAB works, for helping debug existing code, for helping locate MATLAB functions with specific purpose like "weibel fitting". It is not a resource for researching what imafe processing or deep learning algorithms exist and their tradeoffs.

Iniciar sesión para comentar.

Etiquetas

Productos


Versión

R2017a

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by