基于多特征扩散方法的显著性物体检测

叶锋; 洪斯婷; 陈家祯; 郑子华; 刘广海

doi:10.11999/JEIT170827

基于多特征扩散方法的显著性物体检测

doi: 10.11999/JEIT170827

基金项目:

国家自然科学基金(61671077, 61463008)，福建省自然科学基金(2017J01739)，福建省教育厅项目(JA15136)，福建师范大学教学改革研究项目(I201602015)

计量
- 文章访问数: 1414
- HTML全文浏览量: 166
- PDF下载量: 150
- 被引次数: 0
出版历程
- 收稿日期: 2017-08-23
- 修回日期: 2018-01-11
- 刊出日期: 2018-05-19

Salient Object Detection via Multi-feature Diffusion-based Method

Funds:

The National Natural Science Foundation of China (61671077, 61463008), The Natural Science Foundation of Fujian Province (2017J01739), The Scientific Research Fund of Fujian Education Department (JA15136), The Teaching Reform Project of Fujian Normal University (I201602015)

摘要

摘要: 现有的大部分基于扩散理论的显著性物体检测方法只用了图像的底层特征来构造图和扩散矩阵，并且忽视了显著性物体在图像边缘的可能性。针对此，该文提出一种基于图像的多层特征的扩散方法进行显著性物体检测。首先，采用由背景先验、颜色先验、位置先验组成的高层先验方法选取种子节点。其次，将选取的种子节点的显著性信息通过由图像的底层特征构建的扩散矩阵传播到每个节点得到初始显著图，并将其作为图像的中层特征。然后结合图像的高层特征分别构建扩散矩阵，再次运用扩散方法分别获得中层显著图、高层显著图。最后，非线性融合中层显著图和高层显著图得到最终显著图。该算法在3个数据集MSRA10K，DUT-OMRON和ECSSD上，用3种量化评价指标与现有4种流行算法进行实验结果对比，均取得最好的效果。
- 显著性物体检测 /
- 扩散方法 /
- 多层特征 /
- 高层先验
Abstract: Most existing salient object detection methods based on diffusion theory usually only use one feature of image to construct graph and diffusion matrix, and ignore the possibility that salient objects appear at the border regions of the image. In this paper, a diffusion method based on the multi-layer features of image is proposed to detect salient objects. Firstly, the seed nodes are selected by adopting the high-level prior method, which is composed of background prior, color prior, and location prior. Then, the initial saliency map is obtained by propagating the saliency information carried by the selected seed nodes to each nodes via the diffusion matrix constructed by the low-level feature of the image, and used as the middle-level feature of image. The diffusion matrices are re-synthesized again by the middle-level feature and the high-level feature of the image, and the middle-level saliency map and the high-level saliency map are obtained by the diffusion-based method respectively. The final saliency map is obtained by nonlinearly combining the the middle-level and high-level saliency map. Results on three datasets, MSRA10K, DUT-OMRON and ECSSD, show that the proposed method achieves superior performance compared with the four state-of-art methods in terms of three evaluation metrics.
- Salient object detection /
- Diffusion method /
- Multi-layer features /
- High-level prior

HTML全文

参考文献(25)

JERRIPOTHULA K R, CAI J, and YUAN J. Image Co-segmentation via saliency Co-fusion[J]. IEEE Transactions on Multimedia, 2016, 18(9): 1896-1909. doi: 10.1109/TMM.2016.2576283.

LUO P, TIAN Y, WANG X, et al. Switchable deep network for pedestrian detection[C]. IEEE Computer Vision and Pattern Recognition, Columbus, USA, 2014: 899-906. doi: 10.1109/CVPR.2014.120.

ZHAO R, OUYANG W, and WANG X. Unsupervised salience learning for person re-identification[C]. IEEE Computer Vision and Pattern Recognition, Portland, Oregon, USA, 2013: 3586-3593. doi: 10.1109/CVPR.2013.460.

LIU G H, YANG J Y, and LI Z Y. Content-based image retrieval using computational visual attention model[J]. Pattern Recognition, 2015, 48(8): 2554-2566. doi: 10.1016/ j.patcog.2015.02.005

唐红梅, 吴士婧, 郭迎春, 等. 自适应阈值分割与局部背景线索结合的显著性检测[J]. 电子与信息学报, 2017, 39(7): 1592-1598. doi: 10.11999/JEIT160984.

TANG Hongmei, WU Shijing, GUO Yingchun, et al. Saliency detection based on adaptive threshold segmentation and local background clues[J]. Journal of Electronics Information Technology, 2017, 39(7): 1592-1598. doi: 10.11999/JEIT160984.

JIANG H, WANG J, YUAN Z, et al. Automatic salient object segmentation based on context and shape prior[C]. British Machine Vision Conference, Dundee, UK, 2011: 110.1-110.12. doi: 10.5244/C.25.110.

WANG L, WANG L, LU H, et al. Saliency detection with recurrent fully convolutional networks[C]. European Conference on Computer Vision, Amsterdam, Netherlands, 2016: 825-841. doi: 10.1007/978-3-319-46493-0_50.

YANG J and YANG M H. Top-down visual saliency via joint CRF and dictionary learning[J]. IEEE Transactions on Pattern Analysis Machine Intelligence, 2017, 39(3): 576-588. doi: 10.1109/TPAMI.2016.2547384.

HU P and RAMANAN D. Bottom-up and top-down reasoning with hierarchical rectified gaussians[C]. IEEE Conference on Computer Vision and Pattern Recognition, Las Vegas, USA, 2016: 5600-5609. doi: 10.1109/CVPR.2016. 604.

PERAZZI F, Krhenbhl P, PRITCH Y, et al. Saliency filters: Contrast based filtering for salient region detection[C]. IEEE Computer Vision and Pattern Recognition, Rhode Island, 2012: 733-740. doi: 10.1109/CVPR.2012. 6247743.

YAN Q, XU L, SHI J, et al. Hierarchical saliency detection[C]. IEEE Conference on Computer Vision and Pattern Recognition, Portland, USA, 2013: 1155-1162. doi: 10.1109/ CVPR.2013.153.

WANG Q, ZHENG W, and PIRAMUTHU R. GraB: Visual saliency via novel graph model and background priors[C]. IEEE Conference on Computer Vision and Pattern Recognition, Las Vegas, USA, 2016: 535-543. doi: 10.1109/ CVPR.2016.64.

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

HAREL J, KOCH C, and PERONA P. Graph-based visual saliency[C]. International Conference on Neural Information Processing Systems, Vancouver, Canada, 2006: 545-552. doi: 10.1.1.70.2254.

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

YU J G, XIA G S, GAO C, et al. A computational model for object-based visual saliency: Spreading attention along gestalt cues[J]. IEEE Transactions on Multimedia, 2016, 18(2): 273-286. doi: 10.1109/TMM.2015.2505908.

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

JIANG P, VASCONCELOS N, and PENG J. Generic promotion of diffusion-based salient object detection[C]. IEEE International Conference on Computer Vision, Santiago, Chile, 2015: 217-225. doi: 10.1109/ICCV.2015.33.

ACHANTA R, SHAJI A, SMITH K, et al. SLIC superpixels compared to state-of-the-art superpixel methods[J]. IEEE Transactions on Pattern Analysis Machine Intelligence, 2012, 34(11): 2274-2282. doi: 10.1109/TPAMI.2012.120.

LUXBURG U V. A tutorial on spectral clustering[J]. Statistics and Computing, 2007, 17(4): 395-416. doi: 10.1007/ s11222-007-9033-z.

PENG H, LI B, LING H, et al. Salient object detection via structured matrix decomposition[J]. IEEE Transactions on Pattern Analysis Machine Intelligence, 2017, 39(4): 818-832. doi: 10.1109/TPAMI.2016.2562626.

WU Y. A unified approach to salient object detection via low rank matrix recovery[C]. IEEE Computer Vision and Pattern Recognition, Rhode Island, 2012: 853-860. doi: 10.1109/ CVPR.2012.6247758.

ZHANG Lihe, YANG C, LU H, et al. Ranking saliency[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(9): 1892-1904. doi: 10.1109/TPAMI. 2016.2609426.

Borji A, CHENH Mingming, JIANG Huaizu, et al. Salient object detection: A benchmark[C]. IEEE Computer Vision and Pattern Recognition, Boston, USA, 2015: 5706-5722. doi: 10.1109/TIP.2015.2487833.

施引文献

资源附件(0)

访问统计