自然场景中多类目标识别的算法研究

J4 ›› 2012, Vol. 34 ›› Issue (3): 91-95.

自然场景中多类目标识别的算法研究

吴士林1,2,3,4，朱枫1,3,4

(1.中国科学院沈阳自动化研究所,辽宁沈阳 110016；2.中国科学院研究生院,北京100049；
3.中国科学院光电信息处理重点实验室,辽宁沈阳 110016；
4.辽宁省图像理解与视觉计算重点实验室，辽宁沈阳 110016)

收稿日期:2011-05-04 修回日期:2011-08-21 出版日期:2012-03-26 发布日期:2012-03-25

MultiClass Object Recognition in Natural Scenes

WU Shilin1,2,3,4，ZHU Feng1,3,4

（1.Shenyang Institute of Automation,Chinese Academy of Sciences,Shenyang 110016;
2.Graduate School,Chinese Academy of Sciences,Beijing 100049;
3.Key Laboratory of OpticalElectronics Information Processing,
Chinese Academy of Sciences,Shenyang 110016;
4.Key Laboratory of Image Understanding and Computer Vision,Shenyang 110016,China)

Received:2011-05-04 Revised:2011-08-21 Online:2012-03-26 Published:2012-03-25

摘要/Abstract

摘要：

为了实现复杂自然场景中多类目标的识别与分割，本文利用条件概率模型(CM)对目标特征进行建模，融合了纹理特征、纹理环境特征和位置特征，并采用场景类别对各类目标间的相互约束关系进行建模，在此基础上研究基于场景类别的条件概率模型(sCM)在多类目标识别与分割中的应用。本文选用Oliva & Torralba数据库对模型进行实验并与国外其他方法进行了比较。实验结果表明, 该算法在多类目标识别与分割中取得很好的结果，在提高总体识别率的同时提高了物体边缘部分识别与分割的正确率，更有效地提高了视觉效果。

关键词: 目标识别, 多类, 图像分割, Jointboost算法

Abstract:

In this paper, a conditional model (CM) is used to incorporate different feature potentials including texture, texture environment and location features of objects for multiclass object recognition and segmentation in complex natural images. Besides, we model the relationship between different objects by the scene of images and propose a new scenebased conditional model called the sCM model. We investigate the performance of our model in the classbased pixelwise segmentation of images on the Oliva & Torralba database and compare its result with other methods. The results show that our themebased RCRF model significantly improves the accuracy of objects in the whole database. More significantly, a large perceptual improvement is gained, i.e. the details of different objects are correctly labeled.

Key words: object recognition；multiclass；image segmentation；Jointboost

吴士林1,2,3,4，朱枫1,3,4. 自然场景中多类目标识别的算法研究[J]. J4, 2012, 34(3): 91-95.

WU Shilin1,2,3,4，ZHU Feng1,3,4. MultiClass Object Recognition in Natural Scenes[J]. J4, 2012, 34(3): 91-95.

参考文献［13］

［1］	Shotton J, Winn J, Rother C, et al. TextonBoost: Joint Appearance, Shape and Context Modeling for MultiClass Object Recognition and Segmentation［C］∥Proc of European Conference on Computer Vision, 2006, 3951:115.
［2］	Yang L, Meer P, Foran D J. Multiple Class Segmentation Using a Unified Framework over MeanShift Patches［C］∥Proc of IEEE Conference on Computer Vision and Pattern Recognition Minneapolis, 2007:18.
［3］	Schroff F, Criminisi A, Zisserman A. Object Class Segmentation Using Random Forests［C］∥Proc of the 19th British Machine Vision Conference, 2008.
［4］	Gould S, Rodgers J, Cohen D, et al. MultiClass Segmentation with Relative Location Prior［J］.International Journal Computer Vision, 2008, 80(3):300316.
［5］	He X, Zemel R, CarreiraPerpinan M. Multiscale Conditional Random Fields for Image Labeling［C］∥Proc of IEEE Conf on Comput Vision and Pattern Recognition, 2004,2:695702.
［6］	Winn J, Shotton J. The Layout Consistent Random Field for Recognizing and Segmenting Partially Occluded Objects［C］∥Proc of IEEE Conference on Computer Vision and Pattern Recognition, 2006, 1:3744.
［7］	Opelt A, Pinz A, Zisserman A. Incremental Learning of Object Detectors Using a Visual Shape Alphabet［C］∥Proc of IEEE Conference on Computer Vision and Pattern Recognition, 2006, 1:310.
［8］	Winn J, Jojic N. LOCUS: Learning Object Classes with Unsupervised Segmentation［C］∥Proc of International Conference on Computer Vision,2005, 1:756763.
［9］	Kumar P, Torr P, Zisserman A. Obj Cut［C］∥Proc of IEEE Conference on Computer Vision and Pattern Recognition, 2005, 1:1825.
［10］	Winn J, Criminisi A, Minka T. Object Categorization by Learned Universal Visual Dictionary［C］∥Proc of International Conference on Computer Vision, 2005, 2:18001807.
［11］	Sutton C, McCallum A. Piecewise Training of Undirected Models［C］∥Proc of the 21st Conference on Uncertainty in Artifcial Intelligence,2005:568575.
［12］	Torralba A, Murphy K P, Freeman W T. Sharing Visual Features for Multiclass and Multiview Object Detection［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2007, 19(5):854869.
［13］	Johnson M. Semantic Segmentation and Image Search:［Phd Thesis］［D］. University of Cambridge, 2008.

[1]	吴从中, 董浩, 方静. 基于注意力机制的自适应滤波遥感图像分割网络[J]. 计算机工程与科学, 2022, 44(11): 2010-2018.
[2]	何涛, 施慧莉, 李大亮. 基于深度学习的SAR目标识别DSP设计[J]. 计算机工程与科学, 2022, 44(08): 1357-1363.
[3]	仇静博, 燕雪峰, 汪俊, 郭延文, 魏明强, . 基于全卷积神经网络的单幅隧道图像裂纹提取算法[J]. 计算机工程与科学, 2022, 44(05): 845-854.
[4]	曾增峰, 环宇翔, 邹卓, 郑立荣. 多尺度深度监督的高度近视萎缩病变分割方法[J]. 计算机工程与科学, 2021, 43(07): 1264-1272.
[5]	蒋芸, 高静, 王发林. SAG-Net：用于联合视盘和视杯分割的新型跳过注意力指导网络[J]. 计算机工程与科学, 2021, 43(07): 1273-1282.
[6]	张俊鹏, 刘辉, 李清荣. 基于FCN-LSTM的工业烟尘图像分割[J]. 计算机工程与科学, 2021, 43(05): 906-916.
[7]	陈爱莲, 丁正龙, 詹曙. 基于对抗学习和多尺度特征融合的前列腺MR图像分割[J]. 计算机工程与科学, 2021, 43(04): 697-703.
[8]	朱倩倩, 车文刚, 苗晗. 数字化藏文古籍中多样性字体的实现方法研究[J]. 计算机工程与科学, 2020, 42(11): 2073-2079.
[9]	赵仁和，王军锋. 自适应尺度的局部强度聚类图像分割模型[J]. 计算机工程与科学, 2020, 42(06): 1043-1048.
[10]	徐秋平. 基于人机交互式图割的目标快速提取[J]. 计算机工程与科学, 2020, 42(02): 299-306.
[11]	王卫民1,符首夫1,顾榕蓉1,王东升1,何林容2,关文斌3. 基于卷积神经网络的虫情图像分割和计数方法[J]. 计算机工程与科学, 2020, 42(01): 110-116.
[12]	任佳1,2，张胜男1，董超2，赵敏钧1. 基于改进模糊C均值的海面目标图像分割算法[J]. 计算机工程与科学, 2019, 41(05): 858-864.
[13]	吴尚智1，佘志用2，张霞1，赵慧琴3. 利用变精度粗糙熵的图像分割算法[J]. 计算机工程与科学, 2018, 40(10): 1837-1843.
[14]	易三莉，张桂芳，贺建峰，李思洁. 基于最大类间方差的最大熵图像分割[J]. 计算机工程与科学, 2018, 40(10): 1874-1881.
[15]	刘汉强，张青. 区间模糊谱聚类图像分割方法[J]. 计算机工程与科学, 2018, 40(09): 1611-1616.