An efficient video stabilization system for low computational power devices

Autor:	Tae-Shick Wang, Tae-Chan Kim
Rok vydání:	2013
Předmět:	Motion compensation Video post-processing Computer science business.industry ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION Video processing Image stabilization Uncompressed video Video tracking Computer vision Video denoising Artificial intelligence business Block-matching algorithm
Zdroj:	ICCE-Berlin
Popis:	In this paper, we describe an efficient video stabilization (VS) system for low power devices, which produces an unwanted motion compensated video based on feature point (FP) extraction and matching methods. In our system, by extracting the constant number of FPs for every frame, video can be robustly stabilized with the consistent time complexity regardless of scene complexity. Experimental results show that our VS system can efficiently stabilize the video.
Databáze:	OpenAIRE
Externí odkaz:	https://explore.openaire.eu/search/publication?articleId=doi_________::277792dec0a5b5243b706ad566b3a0bc https://doi.org/10.1109/icce-berlin.2013.6698025 Zobrazit plný text záznamu