浏览全部资源
扫码关注微信
重庆大学 光电技术与系统教育部重点实验室,重庆400044
[ "时 旭(1998-),男,河南周口人,硕士研究生,2020年于华侨大学获得学士学位,现为重庆大学光电工程学院研究生,主要从事图像处理及深度学习等方面的研究。E-mail: shixu@cqu.edu.cn" ]
[ "黄 鸿(1980-),男,湖南新宁人,博士生导师,2003、2005、2008年于重庆大学分别获得学士、硕士和博士学位。主要从事医学影像智能化处理、模式识别、深度学习等方面的研究。E-mail: hhuang@cqu.edu.cn" ]
收稿日期:2021-02-27,
修回日期:2021-04-30,
纸质出版日期:2022-04-25
移动端阅览
时旭,李远,黄鸿.面向高光谱显微图像血细胞分类的空-谱可分离卷积神经网络[J].光学精密工程,2022,30(08):960-969.
SHI Xu,LI Yuan,HUANG Hong.Spatial-spectral separable convolutional neural network for cell classification of hyperspectral microscopic images[J].Optics and Precision Engineering,2022,30(08):960-969.
时旭,李远,黄鸿.面向高光谱显微图像血细胞分类的空-谱可分离卷积神经网络[J].光学精密工程,2022,30(08):960-969. DOI: 10.37188/OPE.20223008.0960.
SHI Xu,LI Yuan,HUANG Hong.Spatial-spectral separable convolutional neural network for cell classification of hyperspectral microscopic images[J].Optics and Precision Engineering,2022,30(08):960-969. DOI: 10.37188/OPE.20223008.0960.
深度学习已经在高光谱血细胞图像分类中获得广泛应用。然而,传统深度学习模型需要大量标记数据作为样本,忽略了高光谱图像“图谱合一”的性质,不能充分挖掘高光谱图像内蕴信息,且存在参数多、复杂度高问题。针对上述问题,提出了空-谱可分离卷积神经网络(S
3
CNN),在降低模型复杂度的同时有效提升高光谱血细胞图像分类性能。根据高光谱血细胞图像分布的空间一致性,S
3
CNN模型首先通过空-谱联合距离(SSCD)得到训练集中各像素点的空-谱近邻,并对这些近邻点赋予与相应中心像素点相同的标签,进行样本扩充,然后在网络模型中采用一组深度卷积和点卷积代替经典卷积,优化了模型复杂度,实现血细胞分类。在Bloodcells1-3和Bloodcells2-2两个不同场景下的高光谱血细胞数据集上的实验结果显示,本文所提算法的总体分类精度分别达到87.32%、89.02%。与其他传统血细胞分类算法相比,本文算法能有效提升高光谱血细胞图像的分类性能。在训练时间上,所采用的可分离卷积模型比经典卷积模型减少27%。实验结果表明,所提网络框架不仅能有效提升高光谱血细胞分类性能,且可减少模型训练时间。
In recent years, with the development of computer science,deep learning plays a critical role in the classification of hyperspectral bloodcell images. However, traditional deep learning models require a large amounts of manually annotated training data, and ignore the nature of “graph-spectral uniformity” property of hyperspectral image. As a result, these methods can not explore the intrinsic information of hyperspectral images. In addition, traditional convolutional neural network methods have too many parameters, which takes a great deal of time to be trained. Aiming at these two shortcomings, a spatial-spectral separable convolutional neural network (S
3
CNN) is proposed to improve the classification performance of bloodcell hyperspectral image and reduce the complexity of the model.First, due to the spatial consistency of the hyperspectral bloodcell image distribution, a spatial-spectral combined distance (SSCD) was proposed to select the spatial-spectral nearest neighbor of each pixel and expand the training samples. At the same time, in the following neural network model, a group of depth convolution and point convolution are used to replace classical convolution and optimize the complexity of the model.The experimental result on bloodcell1-3 and bloodcell2-2 datasets show that the overall classification accuracies reaches 87.32% and 89.02%, respectively. Compared with other classification algorithms of bloodcells, the proposed S
3
CNN achieves much higher classification accuracy. The training time of the separable convolution model is 27% less than that of the classical convolution model.Experimental results show that the proposed S
3
CNN is an effective method to improve the classification performance of hyperspectral bloodcell and reduce model training time.
WU Q , LIU J H , MA Q H , et al . White blood cell count as a mediator of the relationship between depressive symptoms and all-cause mortality: a community-based cohort study [J]. Archives of Gerontology and Geriatrics , 2021 , 94 : 104343 . doi: 10.1016/j.archger.2021.104343 http://dx.doi.org/10.1016/j.archger.2021.104343
MA X L , LAN F , ZHANG Y Q . Associations between C-reactive protein and white blood cell count, occurrence of delayed cerebral ischemia and poor outcome following aneurysmal subarachnoid hemorrhage: a systematic review and meta-analysis [J]. Acta Neurologica Belgica , 2021 , 121 ( 5 ): 1311 - 1324 . doi: 10.1007/s13760-020-01496-y http://dx.doi.org/10.1007/s13760-020-01496-y
USTUN C , MORGAN E A , RITZ E M , et al . Core-binding factor acute myeloid leukemia with inv(16): older age and high white blood cell count are risk factors for treatment failure [J]. International Journal of Laboratory Hematology , 2021 , 43 ( 1 ): e19 - e25 . doi: 10.1111/ijlh.13338 http://dx.doi.org/10.1111/ijlh.13338
戴春妮 , 李庆利 , 刘锦高 . 一种基于超光谱的血细胞分类方法 [J]. 计算机应用研究 , 2011 , 28 ( 11 ): 4370 - 4372, 4379 . doi: 10.3969/j.issn.1001-3695.2011.11.100 http://dx.doi.org/10.3969/j.issn.1001-3695.2011.11.100
DAI C N , LI Q L , LIU J G . Blood cell classification method based on hyperspectral data [J]. Application Research of Computers , 2011 , 28 ( 11 ): 4370 - 4372, 4379 . (in Chinese) . doi: 10.3969/j.issn.1001-3695.2011.11.100 http://dx.doi.org/10.3969/j.issn.1001-3695.2011.11.100
刘晶 . 自动化血细胞形态学分析及分类关键技术研究 [D]. 济南 : 山东大学 , 2016 .
LIU J . The Study of Key Techniques of Automated Blood Cell Morphological Analysis and Classification [D]. Jinan : Shandong University , 2016 . (in Chinese)
李航 . 统计学习方法( 2版) [M]. 北京 : 清华大学出版社 , 2019 . doi: 10.7763/ijiet.2015.v5.568 http://dx.doi.org/10.7763/ijiet.2015.v5.568
LI H . Statistical Learning Methods ( The 2 nd Edition ) [M]. Beijing : Tsinghua University Press , 2019 . (in Chinese) . doi: 10.7763/ijiet.2015.v5.568 http://dx.doi.org/10.7763/ijiet.2015.v5.568
郑婷月 , 唐晨 , 雷振坤 . 基于全卷积神经网络的多尺度视网膜血管分割 [J]. 光学学报 , 2019 , 39 ( 2 ): 0211002 . doi: 10.3788/AOS201939.0211002 http://dx.doi.org/10.3788/AOS201939.0211002
ZHENG T Y , TANG C , LEI Z K . Multi-scale retinal vessel segmentation based on fully convolutional neural network [J]. Acta Optica Sinica , 2019 , 39 ( 2 ): 0211002 . (in Chinese) . doi: 10.3788/AOS201939.0211002 http://dx.doi.org/10.3788/AOS201939.0211002
DEY R , LU Z J , HONG Y . Diagnostic classification of lung nodules using 3D neural networks [C]. 2018 IEEE 15th International Symposium on Biomedical Imaging . 47,2018 , Washington, DC, USA . IEEE , 2018 : 774 - 778 . doi: 10.1109/isbi.2018.8363687 http://dx.doi.org/10.1109/isbi.2018.8363687
KHASHMAN A . Investigation of different neural models for blood cell type identification [J]. Neural Computing and Applications , 2012 , 21 ( 6 ): 1177 - 1183 . doi: 10.1007/s00521-010-0476-3 http://dx.doi.org/10.1007/s00521-010-0476-3
FABELO H , ORTEGA S , SZOLNA A , et al . In-vivo hyperspectral human brain image database for brain cancer detection [J]. IEEE Access , 2019 , 7 : 39098 - 39116 . doi: 10.1109/access.2019.2904788 http://dx.doi.org/10.1109/access.2019.2904788
郑欣 , 周梅 , 孙力 , 等 . 基于神经网络的显微高光谱乳腺癌组织图像研究 [J]. 第二军医大学学报 , 2018 , 39 ( 8 ): 886 - 891 .
ZHENG X , ZHOU M , SUN L , et al . Micro-hyperspectral breast cancer tissue image analysis based on neural network [J]. Academic Journal of Second Military Medical University , 2018 , 39 ( 8 ): 886 - 891 . (in Chinese)
TRAJANOVSKI S , SHAN C F , WEIJTMANS P J C , et al . Tongue tumor detection in hyperspectral images using deep learning semantic segmentation [J]. IEEE Transactions on Bio-Medical Engineering , 2021 , 68 ( 4 ): 1330 - 1340 . doi: 10.1109/tbme.2020.3026683 http://dx.doi.org/10.1109/tbme.2020.3026683
HUANG Q , LI W , ZHANG B C , et al . Blood cell classification based on hyperspectral imaging with modulated Gabor and CNN [J]. IEEE Journal of Biomedical and Health Informatics , 2020 , 24 ( 1 ): 160 - 170 . doi: 10.1109/jbhi.2019.2905623 http://dx.doi.org/10.1109/jbhi.2019.2905623
WEI X L , LI W , ZHANG M M , et al . Medical hyperspectral image classification based on end-to-end fusion deep neural network [J]. IEEE Transactions on Instrumentation and Measurement , 2019 , 68 ( 11 ): 4481 - 4492 . doi: 10.1109/tim.2018.2887069 http://dx.doi.org/10.1109/tim.2018.2887069
魏峰 , 何明一 , 梅少辉 . 空间一致性邻域保留嵌入的高光谱数据特征提取 [J]. 红外与激光工程 , 2012 , 41 ( 5 ): 1249 - 1254 . doi: 10.3969/j.issn.1007-2276.2012.05.024 http://dx.doi.org/10.3969/j.issn.1007-2276.2012.05.024
WEI F , HE M Y , MEI S H . Hyperspectral data feature extraction using spatial coherence based neighborhood preserving embedding [J]. Infrared and Laser Engineering , 2012 , 41 ( 5 ): 1249 - 1254 . (in Chinese) . doi: 10.3969/j.issn.1007-2276.2012.05.024 http://dx.doi.org/10.3969/j.issn.1007-2276.2012.05.024
MOHAN A , SAPIRO G , BOSCH E . Spatially coherent nonlinear dimensionality reduction and segmentation of hyperspectral images [J]. IEEE Geoscience and Remote Sensing Letters , 2007 , 4 ( 2 ): 206 - 210 . doi: 10.1109/lgrs.2006.888105 http://dx.doi.org/10.1109/lgrs.2006.888105
黄鸿 , 郑新磊 . 加权空-谱与最近邻分类器相结合的高光谱图像分类 [J]. 光学 精密工程 , 2016 , 24 ( 4 ): 873 - 881 . doi: 10.3788/OPE.20162404.0873 http://dx.doi.org/10.3788/OPE.20162404.0873
HUANG H , ZHENG X L . Hyperspectral image classification with combination of weighted spatial-spectral and KNN [J]. Opt. Precision Eng. , 2016 , 24 ( 4 ): 873 - 881 . (in Chinese) . doi: 10.3788/OPE.20162404.0873 http://dx.doi.org/10.3788/OPE.20162404.0873
VELASCO-FORERO S , MANIAN V . Improving hyperspectral image classification using spatial preprocessing [C]. IEEE Geoscience and Remote Sensing Letters. IEEE , : 297- 301 .
PU H Y , CHEN Z , WANG B , et al . A novel spatial–spectral similarity measure for dimensionality reduction and classification of hyperspectral imagery [J]. IEEE Transactions on Geoscience and Remote Sensing , 2014 , 52 ( 11 ): 7008 - 7022 . doi: 10.1109/tgrs.2014.2306687 http://dx.doi.org/10.1109/tgrs.2014.2306687
TAN K , HU J , LI J , et al . A novel semi-supervised hyperspectral image classification approach based on spatial neighborhood information and classifier combination [J]. ISPRS Journal of Photogrammetry and Remote Sensing , 2015 , 105 : 19 - 29 . doi: 10.1016/j.isprsjprs.2015.03.006 http://dx.doi.org/10.1016/j.isprsjprs.2015.03.006
SZEGEDY C , VANHOUCKE V , IOFFE S , et al . Rethinking the inception architecture for computer vision [C]. 2016 IEEE Conference on Computer Vision and Pattern Recognition . 2730,2016 , Las Vegas, NV, USA . IEEE , 2016 : 2818 - 2826 . doi: 10.1109/cvpr.2016.308 http://dx.doi.org/10.1109/cvpr.2016.308
CHOLLET F . Xception: Deep learning with depthwise separable convolutions [C]. IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, USA : CVPR , 2017 : 1251 - 1258 . doi: 10.1109/cvpr.2017.195 http://dx.doi.org/10.1109/cvpr.2017.195
LI Z Y , HUANG H , DUAN Y L . et al . DLPNet: A deep manifold network for feature extraction of hyperspectral imagery [J]. Neural Networks , 2020 , 129 : 7 - 18 . doi: 10.1016/j.neunet.2020.05.022 http://dx.doi.org/10.1016/j.neunet.2020.05.022
0
浏览量
1143
下载量
3
CSCD
关联资源
相关文章
相关作者
相关机构