基于獨(dú)立分量分析的運(yùn)動(dòng)目標(biāo)檢測(cè)算法中對(duì)通道數(shù)選擇和觀測(cè)向量生成方式的實(shí)驗(yàn)和分析
doi: 10.11999/JEIT140197 cstr: 32379.14.JEIT140197
基金項(xiàng)目:
國(guó)家自然科學(xué)基金(61271352)和安徽省自然科學(xué)基金(1408085 QF125)資助課題
Experiments and Analysis on Observation Vector Generation and Channel Number Selection in Motion Detection Algorithm Based on Independent Component Analysis
-
摘要: 現(xiàn)有基于獨(dú)立分量分析(ICA)的運(yùn)動(dòng)目標(biāo)檢測(cè)算法大多采用單一的觀測(cè)向量生成方式和2通道數(shù)據(jù)進(jìn)行檢測(cè)，使得現(xiàn)有算法難以獲得更加完整精確的目標(biāo)形態(tài)。該文在傳統(tǒng)獨(dú)立分量分析算法的基礎(chǔ)上引入4種不同的觀測(cè)向量生成方式并使用更多通道數(shù)據(jù)進(jìn)行實(shí)驗(yàn)，以此更廣泛地涵蓋運(yùn)動(dòng)目標(biāo)的運(yùn)動(dòng)特性并為提取前景提供更多有效信息，使該算法能有效應(yīng)對(duì)緩慢移動(dòng)和低區(qū)分性目標(biāo)。多場(chǎng)景下的量化實(shí)驗(yàn)分析表明，更多通道數(shù)據(jù)的使用以及4種觀測(cè)向量生成方式的綜合在合理的誤檢率代價(jià)下使算法達(dá)到了更高的檢測(cè)正確率。
-
關(guān)鍵詞:
- 運(yùn)動(dòng)目標(biāo)檢測(cè) /
- 計(jì)算機(jī)視覺 /
- 獨(dú)立分量分析 /
- 通道數(shù)選擇 /
- 觀測(cè)向量生成
Abstract: Most of the existing Independent Component Analysis (ICA) based motion detection algorithms use a single observation vector generation method and two-channel data for motion detection, which make the traditional algorithms unable to obtain a more complete and accurate state of the moving objects. In this paper, four different observation vector generation methods are proposed and larger channel numbers are introduced into traditional ICA. The motion characteristics of the moving objects are covered more widely and more information for foreground extraction is obtained from the multi-channel data. These improvements make ICA be able to deal with indistinguishable and slowly moving objects. The quantitative evaluation from different experiments shows that the multi-channel data and the combination of four observation vector generation methods enable ICA to achieve a better performance with a reasonable cost of tiny increase on false alarms. -
計(jì)量
- 文章訪問數(shù): 2218
- HTML全文瀏覽量: 173
- PDF下載量: 565
- 被引次數(shù): 0
下載:
