基于稀疏表達(dá)的超像素跟蹤算法
doi: 10.11999/JEIT140374 cstr: 32379.14.JEIT140374
基金項(xiàng)目:
國(guó)家自然科學(xué)基金(61273078, 61005032)，中央高?；究蒲袠I(yè)務(wù)費(fèi)(N1106040065032)，國(guó)家科技支撐計(jì)劃項(xiàng)目(2013BAK02B01-02)和遼寧省科技計(jì)劃項(xiàng)目(2013231025)資助課題
Superpixel Tracking Based on Sparse Representation
-
摘要: 該文針對(duì)真實(shí)場(chǎng)景下視頻跟蹤過程中可能出現(xiàn)的目標(biāo)形變、運(yùn)動(dòng)和遮擋等問題，該文分別構(gòu)建了基于超像素局部信息的判別式模型和基于顏色與梯度全局信息的產(chǎn)生式模型，通過兩者的結(jié)合提升了目標(biāo)表觀特征描述的可區(qū)分性和不變性；此外，提出一種基于稀疏主成分分析的更新策略，在更新特征字典的同時(shí)減少其冗余度，在判別式模型的更新階段分別對(duì)每幀圖像獲得的跟蹤結(jié)果進(jìn)行二次判別從而避免漂移現(xiàn)象的發(fā)生。實(shí)驗(yàn)結(jié)果表明，與其它跟蹤算法相比，該算法在應(yīng)對(duì)目標(biāo)姿態(tài)變化、背景干擾以及遮擋等復(fù)雜情況時(shí)具有更好的穩(wěn)定性和魯棒性。
-
關(guān)鍵詞:
- 計(jì)算機(jī)視覺 /
- 目標(biāo)跟蹤 /
- 稀疏表達(dá) /
- 超像素分割 /
- 稀疏主成分分析
Abstract: A novel tracking algorithm is proposed that can work robustly in real-world scenarios, in order to overcome the problems associated with severe changes in pose, motion and occlusion. A discriminative model based on the superpixels and a generative model based on global color and gradient features are constructed respectively. Through combining these two models, the distinguishing and invariance of target appearance features description are increased. Furthermore, an update strategy based on sparse principal component analysis is proposed, which can reduce the redundancy of feature dictionary when it updates. A discrimination mechanism is added in the update process of discriminative model to alleviate the drift problem. The experimental results demonstrate that the proposed algorithm performs more stable and robustly compared with several state-of-the-art algorithms when dealing with complex situations such as pose variation, background interference, and occlusion. -
計(jì)量
- 文章訪問數(shù): 2333
- HTML全文瀏覽量: 237
- PDF下載量: 3142
- 被引次數(shù): 0
下載:
