基于显著度融合的自适应分块行人再识别

陈鸿昶; 陈雷; 李邵梅; 朱俊光

doi:10.11999/JEIT170162

基于显著度融合的自适应分块行人再识别

doi: 10.11999/JEIT170162

基金项目:

国家自然科学基金(61379151, 61521003)，河南省杰出青年基金(144100510001)

计量
- 文章访问数: 1355
- HTML全文浏览量: 137
- PDF下载量: 318
- 被引次数: 0
出版历程
- 收稿日期: 2017-02-24
- 修回日期: 2017-04-27
- 刊出日期: 2017-11-19

Person Re-identification of Adaptive Blocks Based on Saliency Fusion

Funds:

The National Natural Science Foundation of China (61379151, 61521003), Outstanding Youth Foundation of Henan Province (144100510001)

摘要

摘要: 针对基于分块匹配的行人再识别中对分块的规则和大小缺乏指导，以及不同分块间的区分度差异问题，该文提出基于显著度融合的自适应分块行人再识别方法。首先，利用启发式思想确定初始聚类中心，并根据图像内容自动确定分块的大小和数目。然后，利用归一化部分曲线下面积计算各块的图像间显著度，利用结构化支持向量机学习各块的图像内显著度，并融合两类显著度得到各块的权重作为匹配得分融合的依据。实验证明，在常用的行人再识别数据集上，该方法能取得较好的识别结果。
- 行人再识别 /
- 分块匹配 /
- 自适应分块 /
- 启发式聚类 /
- 显著度融合
Abstract: In this paper, an adaptive block person re-identification method based on saliency fusion is proposed to solve the problems of the lack of guidance on the rule and size of block in the block matching-based person re-identification, and the differentiation degree between different blocks. Firstly, the heuristic idea is used to determine the initial clustering center, and the size and number of blocks are determined automatically according to the image content. Then, the intra-image salience of each block is calculated using the Area Under the normalized partial Curve (pAUC), the intra-image salience of each block is learned by structured SVM, and the weights of each block are fused as the base of matching Score fusion. Experiments show that this method can achieve better recognition results on the commonly used person re-identification data sets.
- Person re-identification /
- Block matching /
- Adaptive blocks /
- Heuristic clustering /
- Saliency fusion

HTML全文

参考文献(22)

ZHENG W S, GONG S, and XIANG T. Reidentification by relative distance comparison[J]. IEEE Transactions on Pattern Analysis Machine Intelligence, 2013, 35(3): 653-668. doi: 10.1109/TPAMI.2012.138.

ZHANG L, KALASHNIKOV D V, MEHROTRA S, et al. Context-based person identification framework for smart video surveillance[J]. Machine Vision Applications, 2014, 25(7): 1711-1725. doi: 10.1007/s00138-013-0535-8.

HE K, ZHANG X, REN S, et al. Spatial pyramid pooling in deep convolutional networks for visual recognition[J]. IEEE Transactions on Pattern Analysis Machine Intelligence, 2015, 37(9): 1904-1916. doi: 10.1109/TPAMI.2015.2389824.

齐美彬, 檀胜顺, 王运侠, 等. 基于多特征子空间与核学习的行人再识别[J]. 自动化学报, 2016, 33(2): 299-308. doi: 10.16383/j.aas.2016.c150344.

QI Meibin, TAN Shengshun, WANG Yunxia, et al. Multi- feature subspace and kernel learning for pedestrian re- identification[J]. Automation Journal, 2016, 33(2): 299-308. doi: 10.16383/j.aas.2016.c150344.

曾明勇, 吴泽民, 田畅, 等. 基于外观统计特征融合的人体目标再识别[J]. 电子与信息学报, 2014, 36(8): 1844-1851. doi: 10.3724/SP.J.1146.2013.01389.

ZENG Mingyong, WU Zemin, TIAN Chang, et al. Fusing appearance statistical features for person re-identification[J]. Journal of Electronics Information Technology, 2014, 36(8): 1844-1851. doi: 10.3724/SP.J.1146.2013.01389.

FARENZENA M, BAZZANI L, PERINA A, et al. Person re-identification by symmetry-driven accumulation of local features[C]. IEEE Computer Vision and Pattern Recognition, San Francisco, California, USA, 2010: 2360-2367. doi: 10. 1109/CVPR.2010.5539926.

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

ZHAO R, OUYANG W and WANG X. Learning mid-level filters for person re-identification[C]. IEEE Conference on Computer Vision and Pattern Recognition, Columbus, OH, USA, 2014: 144-151. doi: 10.1109/ CVPR.2014.26.

ZHANG L, LI K, ZHANG Y, et al. Adaptive image segmentation based on color clustering for person re-identification[J]. Soft Computing, 2016, 36(2): 1-11. doi: 10.1007/s00500-016-2150-x.

ENDRES I and HOIEM D. Category-independent object proposals with diverse ranking[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2014, 36(2): 222-234. doi: 10.1109/ TPAMI.2013.122.

ALEXE B, DESELAERS T, and FERRARI V. What is an object?[C]. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), San Francisco, California, USA, 2010: 73-80. doi: 10.1109/CVPR. 2010.5540226.

JIANG F, JIA L, SHENG X, et al. Manifold regularization in structured output space for semi-supervised structured output prediction[J]. Neural Computing Applications, 2016, 27(8): 2605-2614. doi: 10.1007/s00521-015-2029-2.

JOACHIMS T, FINLEY T, and YU C N J. Cutting-plane training of structural SVMs[J]. Machine Learning, 2009, 77(1): 27-59. doi: 10.1007/s10994-009-5108-8.

LI Wei and WANG Xiaogang. Locally aligned feature transforms across views[C]. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Columbus, Ohio, USA, 2013: 3594-3601. doi: 10.1109/CVPR. 2013.461.

SCHWARTZ W R and DAVIS L S. Learning discriminative appearance-based models using partial least squares[C]. Proceedings of the XXII Brazilian Symposium on Computer Graphics and Image Processing, Rio De Janeiro, Brazil, 2009: 322-329. doi: 10.1109/SIBGRAPI.2009.42.

DING S, LIN L, WANG G, et al. Deep feature learning with relative distance comparison for person re-identification[J]. Pattern Recognition, 2015, 48(10): 2993-3003. doi: 10.1016/j. patcog.2015.04.005.

陈莹, 霍中花. 多方向显著性权值学习的行人再识别[J]. 中国图象图形学报, 2015, 20(12): 1674-1683. doi: 10.11834/ jig.20151212.

CHEN Ying and HUO Zhonghua. Person re-identification based on multi-directional saliency metric learning[J]. Journal of Image and Graphics, 2015, 20(12): 1674-1683. doi: 10.11834/jig.20151212.

MARTIN Hirzer, PETER M Roth, MARTIN Kostinger, et al. Relaxed pairwise learned metric for person re-identification [C]. Proceedings of the IEEE European Conference on Computer Vision, Firenze, Italy, 2012: 780-793. doi: 10.1007/ 978-3-642-33783-3_56.

DONG Yi, ZHEN Lei, LIAO Shengcai, et al. Deep metric learning for person re-identification[C]. Proceedings of International Conference on Pattern Recognition, Sweden, 2014: 34-39. doi: 10.1109/ICPR.2014.16.

施引文献

资源附件(0)

访问统计