Ubrzavanje kodiranja videa koristeći izvorne slike u Bayerovom formatu

master's thesis

• Title (english): Acceleration of video encoding using source images in Bayer format
• Author: Dominik Pavel
• Mentor: Daniel Hofman (mentor)
• Committee member: Daniel Hofman (član povjerenstva)
• Committee member: Hrvoje Mlinarić (član povjerenstva)
• Committee member: Josip Knezović (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2024-07-02, Croatia

Abstract

Ovaj rad temelji se na obradi slike dobivene s ASI ZWO kamere koristeći ASI_SDK biblioteku, fokusirajući se na dva različita načina enkodiranja i dekodiranja: normalno enkodiranje i maskirano enkodiranje. Normalno enkodiranje uključuje proces dohvaćanja slike s kamere u RGGB obliku, zatim izvlačenje crvenog, zelenog i plavog kanala iz slike kao i proces spajanja tih kanala u YUV444p format, enkodiranje u MP4 format pa zatim dekodiranje nazad u YUV444p format. Za razliku od normalnog, maskirano enkodiranje koristi dodatne korake koji uključuju pretvaranje u YUV422p format prije enkodiranja, maskiranje zelenog kanala kao Y, plavog kanala kao U te crvenog kanala kao V u YUV formatu. Nakon dekodiranja maskirani video ponovno vrši ekstrakciju zelenog, plavog i crvenog kanala i spaja ih u konačnu sliku u YUV444p formatu. Nakon što oba videa prođu proces enkodiranja i dekodiranja, konačni videi uspoređuju se uz pomoć FFmpeg-ovog PSNR-a za dobivanje razlike u kvaliteti videa.

Abstract (english)

This thesis is based on the processing of an image obtained from an ASI ZWO camera using the ASI_SDK library, focusing on two different encoding and decoding methods: normal encoding and masked encoding. Normal encoding involves the process of retrieving the image from the camera in RGGB format, then extracting the red, green and blue channels from the image as well as the process of combining those channels into YUV444p format, encoding into MP4 format and then decoding back into YUV444p format. Unlike normal, masked encoding uses additional steps that include converting to YUV422p format before encoding, masking the green channel as Y, the blue channel as U, and the red channel as V in YUV format. After decoding, the masked video re-extracts the green, blue and red channels and merges them into the final image in YUV444p format. After both videos go through the process of encoding and decoding, the final videos are compared using FFmpeg's PSNR to get the difference in video quality.

• Keywords: ASI ZWO kamera; C++ programski jezik; ASI_SDK biblioteka; OpenCV; FFmpeg; Video enkoder; Video dekoder; Bayerov format; YUV format; Demosaik; Pretvorba prostora boja
• Keywords (english): ASI ZWO camera; C++ programming language; ASI_SDK library; OpenCV; FFmpeg; Video encoder; Video decoder; Bayer's format; YUV format; Demosaic; Color space transformation
• Language: croatian
• URN:NBN: urn:nbn:hr:168:541881
• Study programme: Title: Computing; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka računarstva (magistar/magistra inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2024-10-12 21:46:26