Super Resolution Video Group - PowerPoint PPT Presentation

Super Resolution для сжатого видео Моисейцев Алексей Video Group CS MSU Graphics & Media Lab

Only for Maxus  Содержание  Введение  Example-based SR  SRME  HMRF SR 2 CS MSU Graphics & Media Lab (Video Group)

Only for Maxus  Введение 3 CS MSU Graphics & Media Lab (Video Group)

Only for Maxus  Введение Lossless 2Mbps 1Mbps 256Kbps Bi-cubic source Super-res 4 Super-Resolving Compressed Video with Large Artifacts, Wen-Yi Zhao, ICPR CS MSU Graphics & Media Lab (Video Group) 2004

Only for Maxus  Введение Алгоритмы SR:  Iterative Backprojection (IBP)  Projection Onto Convex Sets (POCS)  Probabilistic Methods  Maximum a posteriori (MAP)  Model-based approach (MBSR)  Example-based 6 CS MSU Graphics & Media Lab (Video Group)

Only for Maxus  Введение IBP POCS C 2 C 1 C f=Pg P 1 g P 2 g C 3 P 3 g g 7 CS MSU Graphics & Media Lab (Video Group)

Only for Maxus  Введение 8 Super resolution: an overview, C Papathanassiou and M Petrou, Geoscience CS MSU Graphics & Media Lab (Video Group) and Remote Sensing Symposium, 2005

Only for Maxus  Введение Типичная скорость работы : 0.15-0.4 fps 9 CS MSU Graphics & Media Lab (Video Group)

Only for Maxus  Содержание  Введение  Example-based SR  SRME  HMRF SR 10 10 10 CS MSU Graphics & Media Lab (Video Group)

Only for Maxus  Example-based SR  В SR только по LR- кадрам существует теоретический предел качества  Иногда есть доступ и к отдельным кадрам в высоком разрешении 11 Resolution enhancement of low quality videos using a high-resolution frame, CS MSU Graphics & Media Lab (Video Group) Tuan Phama, Lucas van Vlieta, Klamer Schutte, SPIE 2006

Only for Example-based SR Maxus  Registration Для компенсации движения используется алгоритм Lucas-Kanade: 12 Resolution enhancement of low quality videos using a high-resolution frame, CS MSU Graphics & Media Lab (Video Group) Tuan Phama, Lucas van Vlieta, Klamer Schutte, SPIE 2006

Only for Example-based SR Maxus  Registration В 99% случаев ошибка не превышает 0.15 LR- пикселя и 0.45 HR- пикселя 13 Resolution enhancement of low quality videos using a high-resolution frame, CS MSU Graphics & Media Lab (Video Group) Tuan Phama, Lucas van Vlieta, Klamer Schutte, SPIE 2006

Only for Example-based SR Maxus  Построение пар HR-LR  Используются LR- блоки 8x8  Рассматриваются повороты блоков  Ключ — 10 AC- коэффициентов LR- блока и 68 граничных пикселей HR- блока 14 Resolution enhancement of low quality videos using a high-resolution frame, CS MSU Graphics & Media Lab (Video Group) Tuan Phama, Lucas van Vlieta, Klamer Schutte, SPIE 2006

Only for Example-based SR Maxus  Восстановление 15 Resolution enhancement of low quality videos using a high-resolution frame, CS MSU Graphics & Media Lab (Video Group) Tuan Phama, Lucas van Vlieta, Klamer Schutte, SPIE 2006

Only for Example-based SR Maxus  Восстановление 16 Resolution enhancement of low quality videos using a high-resolution frame, CS MSU Graphics & Media Lab (Video Group) Tuan Phama, Lucas van Vlieta, Klamer Schutte, SPIE 2006

Only for Example-based SR Maxus  Результат source, frame 0 compressed frame 20 SR, frame 20 MJPEG, Q=50  Обучение на первом кадре  Восстановление двадцатого кадра  17 Resolution enhancement of low quality videos using a high-resolution frame, CS MSU Graphics & Media Lab (Video Group) Tuan Phama, Lucas van Vlieta, Klamer Schutte, SPIE 2006

Only for Example-based SR Maxus  Результат compressed frame 20 SR, frame 20 Spatial example-based SR, frame 20 18 Resolution enhancement of low quality videos using a high-resolution frame, CS MSU Graphics & Media Lab (Video Group) Tuan Phama, Lucas van Vlieta, Klamer Schutte, SPIE 2006

Only for Maxus  Key-frame based SR Идея: использовать HR ключевые кадры 19 Super Resolution of Video Using Key Frames, Fernanda Brandi, Ricardo de CS MSU Graphics & Media Lab (Video Group) Queiroz, Debargha Mukherjee, ISCAS 2008

Only for Key-frame based SR Maxus  Восстановление 20 Super Resolution of Video Using Key Frames, Fernanda Brandi, Ricardo de CS MSU Graphics & Media Lab (Video Group) Queiroz, Debargha Mukherjee, ISCAS 2008

Only for Key-frame based SR Maxus  Результаты 21 Super Resolution of Video Using Key Frames, Fernanda Brandi, Ricardo de CS MSU Graphics & Media Lab (Video Group) Queiroz, Debargha Mukherjee, ISCAS 2008

Only for Key-frame based SR Maxus  Результаты 22 Super Resolution of Video Using Key Frames, Fernanda Brandi, Ricardo de CS MSU Graphics & Media Lab (Video Group) Queiroz, Debargha Mukherjee, ISCAS 2008

Only for Key-frame based SR Maxus  Результаты 23 Super Resolution of Video Using Key Frames, Fernanda Brandi, Ricardo de CS MSU Graphics & Media Lab (Video Group) Queiroz, Debargha Mukherjee, ISCAS 2008

Only for Maxus  Example-based SR  Слабая обоснованность метода  Требуется наличие специфичного видеопотока  Зависимость от обучающей выборки 24 CS MSU Graphics & Media Lab (Video Group)

Only for Maxus  Содержание  Введение  Example-based SR  SRME  HMRF SR 25 25 25 CS MSU Graphics & Media Lab (Video Group)

Only for Maxus  SRME Simultaneous motion estimation and resolution enhancement of 26 compressed low resolution video. CS MSU Graphics & Media Lab (Video Group) Javier Mateos, A. K. Katsaggelos, Rafael Molina, ICIP, 2000

Only for Maxus  SRME Simultaneous motion estimation and resolution enhancement of 27 compressed low resolution video. CS MSU Graphics & Media Lab (Video Group) Javier Mateos, A. K. Katsaggelos, Rafael Molina, ICIP, 2000

Only for Maxus  SRME — апостериорная вероятность — интересующее решение Simultaneous motion estimation and resolution enhancement of 28 compressed low resolution video. CS MSU Graphics & Media Lab (Video Group) Javier Mateos, A. K. Katsaggelos, Rafael Molina, ICIP, 2000

Only for SRME Maxus  Задание вероятностей (1) Simultaneous motion estimation and resolution enhancement of 29 compressed low resolution video. CS MSU Graphics & Media Lab (Video Group) Javier Mateos, A. K. Katsaggelos, Rafael Molina, ICIP, 2000

Only for SRME Maxus  Задание вероятностей (2) Simultaneous motion estimation and resolution enhancement of 30 compressed low resolution video. CS MSU Graphics & Media Lab (Video Group) Javier Mateos, A. K. Katsaggelos, Rafael Molina, ICIP, 2000

Only for SRME Maxus  Задание вероятностей (3) Simultaneous motion estimation and resolution enhancement of 31 compressed low resolution video. CS MSU Graphics & Media Lab (Video Group) Javier Mateos, A. K. Katsaggelos, Rafael Molina, ICIP, 2000

Only for SRME Maxus  Решение (1) Simultaneous motion estimation and resolution enhancement of 32 compressed low resolution video. CS MSU Graphics & Media Lab (Video Group) Javier Mateos, A. K. Katsaggelos, Rafael Molina, ICIP, 2000

Only for SRME Maxus  Решение (2) Две итерации:  Компенсация движения  Построение улучшенного кадра Simultaneous motion estimation and resolution enhancement of 33 compressed low resolution video. CS MSU Graphics & Media Lab (Video Group) Javier Mateos, A. K. Katsaggelos, Rafael Molina, ICIP, 2000

Only for Maxus  SRME source Y-PSNR = 20.44dB Simultaneous motion estimation and resolution enhancement of 34 compressed low resolution video. CS MSU Graphics & Media Lab (Video Group) Javier Mateos, A. K. Katsaggelos, Rafael Molina, ICIP, 2000

Only for Maxus  SRME bilinear Y-PSNR = 21.98dB Simultaneous motion estimation and resolution enhancement of 35 compressed low resolution video. CS MSU Graphics & Media Lab (Video Group) Javier Mateos, A. K. Katsaggelos, Rafael Molina, ICIP, 2000

Only for Maxus  SRME proposed SR Y-PSNR = 25.64dB Simultaneous motion estimation and resolution enhancement of 36 compressed low resolution video. CS MSU Graphics & Media Lab (Video Group) Javier Mateos, A. K. Katsaggelos, Rafael Molina, ICIP, 2000

Only for Maxus  Содержание  Введение  Example-based SR  SRME  HMRF SR 37 37 37 CS MSU Graphics & Media Lab (Video Group)