融合顯著深度特征的RGB-D圖像顯著目標(biāo)檢測(cè)

吳建國(guó); 邵婷; 劉政怡*

doi:10.11999/JEIT161304

融合顯著深度特征的RGB-D圖像顯著目標(biāo)檢測(cè)

doi: 10.11999/JEIT161304

基金項(xiàng)目:

國(guó)家科技支撐計(jì)劃(2015BAK24B00)，高等學(xué)校博士學(xué)科點(diǎn)專項(xiàng)科研基金(20133401110009)，安徽高校省級(jí)自然科學(xué)研究項(xiàng)目(KJ2015A009)，安徽大學(xué)信息保障技術(shù)協(xié)同創(chuàng)新中心開放課題

計(jì)量
- 文章訪問數(shù): 1561
- HTML全文瀏覽量: 211
- PDF下載量: 463
- 被引次數(shù): 0
出版歷程
- 收稿日期: 2016-12-08
- 修回日期: 2017-05-22
- 刊出日期: 2017-09-19

RGB-D Saliency Detection Based on Integration Feature of Color and Depth Saliency Map

Funds:

The National Key Technology RD Program (2015BAK24B00), The Specialized Research Fund for the Doctoral Program of Higher Education of China (20133401110009), Key Program of Natural Science Project of Educational Commission of Anhui Province (KJ2015A009), Open Funds of Co-Innovation Center for Information Supply Assurance Technology of Anhui University

摘要

摘要: 深度信息被證明是人類視覺的重要組成部分，然而大部分顯著性檢測(cè)工作側(cè)重于2維圖像上的方法，并不能很好地利用深度進(jìn)行RGB-D圖像顯著性檢測(cè)。該文提出一種融合顯著深度特征的RGB-D圖像顯著目標(biāo)檢測(cè)方法，提取基于顏色和深度顯著圖的綜合特征，根據(jù)構(gòu)圖先驗(yàn)和背景先驗(yàn)的方法進(jìn)行顯著目標(biāo)檢測(cè)。首先，對(duì)原始深度圖進(jìn)行預(yù)處理：使用背景頂點(diǎn)區(qū)域、構(gòu)圖交點(diǎn)和緊密度處理深度圖，多角度融合形成深度顯著圖，并作為顯著深度特征，結(jié)合顏色特征形成綜合特征；其次，從前景角度，將綜合特征通過邊連接權(quán)重構(gòu)造關(guān)聯(lián)矩陣，根據(jù)構(gòu)圖先驗(yàn)，假設(shè)多層中心矩形為前景種子，通過流形排序方法計(jì)算出RGB-D圖像的前景顯著圖；從背景角度，根據(jù)背景先驗(yàn)以及邊界連通性計(jì)算出背景顯著圖；最后，將前景顯著圖和背景顯著圖進(jìn)行融合并優(yōu)化得到最終顯著圖。實(shí)驗(yàn)采用RGB-D1000數(shù)據(jù)集進(jìn)行顯著性檢測(cè)，并與4種不同的方法進(jìn)行對(duì)比，所提方法的顯著性檢測(cè)結(jié)果更接近人工標(biāo)定結(jié)果，PR(查準(zhǔn)率-查全率)曲線顯示在相同召回率下準(zhǔn)確率高于其他方法。
- 顯著目標(biāo)檢測(cè) /
- 顯著深度特征 /
- 多層中心矩形 /
- 流形排序 /
- 構(gòu)圖先驗(yàn) /
- 背景先驗(yàn)
Abstract: Depth information is proved to be an important part of human vision. However, most saliency detection methods based on 2D images do not make good use of depth information, thus an effective saliency detection method for RGB-D image is presented. It extracts color feature combined with depth saliency feature and detects salient objects based on photographic composition prior and background prior. First, original depth map is preprocessed to form depth saliency feature by background vertex area, photographic composition intersections, and compactness method. Then the association matrix is constructed by the adjacency weight of comprehensive feature. Manifold ranking is running from foreground view to form foreground saliency map based on photographic composition prior and fusion of depth saliency feature and color feature. In order to correct the error caused by assumption, the boundary connectivity is used to suppress background from background view. Final saliency map builds on fusion of foreground and background saliency map. Experiments compared with 4 different methods on RGB-D1000 database show that the proposed method has better precision-recall curve and outperforms the state- of-the-art methods.
- Salient object detection /
- Depth saliency map /
- Multi-layer center rectangle /
- Manifold ranking /
- Photographic composition prior /
- Background prior

HTML全文

參考文獻(xiàn)(23)

DING Y, XIAO J, and YU J. Importance filtering for image retargeting[C]. IEEE Conference on Computer Vision Pattern Recognition. Colorado Springs, CO, USA, 2011: 89-96. doi: 10.1109/CVPR.2011.5995445.

SIAGIAN C and ITTI L. Rapid biologically-inspired scene classification using features shared with visual attention[J]. IEEE Transactions on Pattern Analysis Machine Intelligence, 2007, 29(2): 300-312. doi: 10.1109/TPAMI.2007. 40.

MAHADEVAN V and VASCONCELOS N. Saliency-based discriminant tracking[C]. IEEE Computer Society Conference on Computer Vision and Pattern Recognition. Miami, FL, USA, 2009: 1007-1013. doi: 10.1109/CVPR.2009. 5206573.

ROTHER C, KOLMOGOROV V, and BLAKE A. GrabCut: Interactive foreground extraction using iterated graph cuts[J]. ACM Transactions on Graphics, 2004, 23(3): 309-314. doi: 10.1109/CGIV.2013.34.

LIN W Y, WU P C, and CHEN B R. Image retargeting using depth enhanced saliency[J]. Proceedings of 3DSA, 2013, 7:1.

FENG D, BARNES N, YOU S, et al. Local background enclosure for RGB-D salient object detection[C]. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas, NV, USA, 2016: 2343-2350. doi: 10. 1109/CVPR.2016.257.

CHENG Y, FU H, WEI X, et al. Depth enhanced saliency detection method[J]. European Journal of Histochemistry, 2014, 55(1): 301-308. doi: 10.1145/2632856.2632866.

REN J, GONG X, YU L, et al. Exploiting global priors for RGB-D saliency detection[C]. 2015 IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW). Boston, MA, USA, 2015: 25-32. doi: 10.1109/ CVPRW.2015.7301391.

黃子超, 劉政怡. 特征融合與S-D概率矯正的RGB-D顯著檢測(cè)[J]. 中國(guó)圖象圖形學(xué)報(bào), 2016, 21(10): 1392-1401.

HUANG Zichao and LIU Zhengyi. Feature integration and S-D probability correction based RGB-D saliency detection[J]. Journal of Image and Graphics, 2016, 21(10): 1392-1401.

GOFERMAN S, ZELNIK-MANOR L, and TAL A. Context- aware saliency detection[J]. IEEE Transactions on Pattern Analysis Machine Intelligence, 2012, 34(10): 1915-1926. doi: 10.1109/TPAMI.2011.272.

王嬌嬌, 劉政怡. 多尺度構(gòu)圖先驗(yàn)的顯著目標(biāo)檢測(cè)[J]. 中國(guó)圖象圖形學(xué)報(bào), 2015, 20(12): 1664-1673.

WANG Jiaojiao and LIU Zhengyi. Multi-scale saliency detection based on composition prior[J]. Journal of Image and Graphics, 2015, 20(12): 1664-1673.

ZHOU L, YANG Z, YUAN Q, et al. Salient region detection via integrating diffusion-based compactness and local contrast[J]. IEEE Transactions on Image Processing, 2015, 24(11): 3308-3320. doi: 10.1109/TIP.2015.2438546.

TANG Shuai, WANG Xiaoyu, LU Xutao, et al. Histogram of oriented normal vectors for object recognition with a depth sensor[C]. Asian Conference on Computer Vision 2012. Daejeon, Korea, 2012: 525-538. doi: 10.1007/978-3-642 -37444-9_41.

XUE H, GU Y, LI Y, et al. RGB-D saliency detection via mutual guided manifold ranking[C]. IEEE International Conference on Image Processing. Quebec City, QC, Canada, 2015: 666-670. doi: 10.1109/ICIP.2015.7350882.

YANG C, ZHANG L, LU H, et al. Saliency detection via graph-based manifold ranking[C]. IEEE Conference on Computer Vision and Pattern Recognition. IEEE Computer Society, 2013: 3166-3173. doi: 10.1109/CVPR.2013.407.

ZHU W, LIANG S, WEI Y, et al. Saliency optimization from robust background detection[C]. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Columbus, OH, USA, 2014: 2814-2821. doi: 10.1109/CVPR. 2014.360.

QIN Y, LU H, XU Y, et al. Saliency detection via cellular automata[C]. IEEE Conference on Computer Vision and Pattern Recognition. Boston, MA, USA, 2015: 110-119. doi: 10.1109/CVPR.2015.7298606.

JIANG B, ZHANG L, LU H, et al. Saliency detection via absorbing Markov chain[C]. Proceedings of the IEEE International Conference on Computer Vision. Sydney, NSW, Australia, 2013: 1665-1672. doi: 10.1109/ICCV.2013.209.

PENG H, LI B, XIONG W, et al. RGBD salient object detection: A benchmark and algorithms[C]. European Conference on Computer Vision. Zurich, Switzerland, 2014: 92-109. doi: 10.1007/978-3-319-10578-9_7.

JU R, GE L, GENG W, et al. Depth saliency based on anisotropic center-surround difference[C]. 2014 IEEE International Conference on Image Processing. Paris, France, 2014: 1115-1119. doi: 10.1109/ICIP.2014.7025222.

WEI Y, WEN F, ZHU W, et al. Geodesic saliency using background priors[C]. European Conference on Computer Vision. Florence, Italy, 2012: 29-42. doi: 10.1007/978-3-642- 33712-3_3.

相關(guān)文章

施引文獻(xiàn)

資源附件(0)

訪問統(tǒng)計(jì)