針對(duì)室內(nèi)安全監(jiān)控的場(chǎng)景特點(diǎn),提出一種穩(wěn)健的運(yùn)動(dòng)目標(biāo)檢測(cè)方法,對(duì)每個(gè)背景像素使
用多個(gè)矢量表示,利用當(dāng)前幀和背景幀像素的色度分量差分進(jìn)行運(yùn)動(dòng)檢測(cè),并選擇性地更新背景模型。實(shí)驗(yàn)表明,該方法能有效消除背景中振動(dòng)對(duì)象的干擾,提高運(yùn)動(dòng)目標(biāo)檢測(cè)的準(zhǔn)確性,有利于消除由于虛假目標(biāo)而導(dǎo)致的誤報(bào)警。
關(guān)鍵字:視頻序列;運(yùn)動(dòng)目標(biāo)檢測(cè);背景差分;背景更新
Abstract: According to the characteristic of the indoor surveillance, this paper presents a robust method for moving object detection, which gives multi vectors for each background pixel, and detects moving object via the chroma difference of the current and background image, at the same time the background model can be selectively updated. The experiments show this method can eliminate the swaying objects in the background and improve the detection of moving objects and reject false alarms.
Keywords: video sequences; moving object detection; background subtraction; background
update