摘  要：在小波變換理論基礎(chǔ)上，提出對(duì)視頻的運(yùn)動(dòng)估計(jì)量進(jìn)行處理來(lái)達(dá)到視頻去噪的方法。首先使用運(yùn)動(dòng)域細(xì)化技術(shù)和全新的基于運(yùn)動(dòng)補(bǔ)償?shù)臅r(shí)間域?yàn)V波器，解決含有虛假運(yùn)動(dòng)矢量的實(shí)時(shí)視頻去噪問(wèn)題，提高對(duì)動(dòng)態(tài)目標(biāo)追蹤的準(zhǔn)確性，增強(qiáng)魯棒性；并且改進(jìn)空間域?yàn)V波器，降低處理的復(fù)雜度。實(shí)驗(yàn)結(jié)果表明：該方法是一種較為簡(jiǎn)單且高效的視頻去噪方法。
關(guān)鍵詞：圖像增強(qiáng)；運(yùn)動(dòng)估計(jì)；視頻編碼；小波變換
中圖分類號(hào)：TP312；TP391.41；TN949.6+3；TM930.12     文獻(xiàn)標(biāo)志碼：A    文章編號(hào)：1674-5124（2013）05-0001-05
Video denoising method based on motion estimation combined with
wavelet-domain and motion-compensated
FENG Chang-jiang， MAO Bo， XUE Bing
（Ordance Engineering College，Shijiazhuang 050003，China）
Abstract: On the basis of wavelet transform theory， a novel video denoising method is presented in this paper， which reuses motion estimation resources from the video coding module. Firstly， it proposed a novel motion-field filtering step that refines the accuracy of the motion estimates to a degree that is required for denoising. Secondly， a novel temporal filter was proposed that is robust against errors in the estimated motion field. Then spatial domain filter was improved to reduce the complexity of the processing. The experimental results show that the proposed method is a relatively simple and efficient in video denoising.
Key words: image enhancement； motion estimation； video coding； wavelet transform