深度语义分割MRF模型的海洋筏式养殖信息提取

doi:10.6040/j.issn.1672-3961.0.2021.300

摘要/Abstract

摘要： 采用无监督方法与深度学习模型结合,解决筏式养殖边缘信息精确提取问题,提出深度语义分割(semantic segmentation, SegNet)-马尔科夫随机场(Markov random field, MRF)模型,该模型提取目标空间细节信息和深度判别特征信息。通过SegNet编码器的卷积和最大池化提取筏式养殖的特征信息和扩大感受野,抑制噪声、误判等现象的产生,模型后端接入MRF模型,计算像素空间领域内的特征信息进行聚类分析来获取目标低水平的空间细节信息,在深度特征信息的基础上较大程度的保留空间特征信息,完善边缘信息并抑制连通区域的产生。试验结果表明,该模型极大减少了特征信息丢失和因海水背景而产生的误判,其分类精度高于95%,明显优于经典无监督算法和单一的深度学习模型。

关键词: 筏式养殖, 卷积神经网络, 深度学习, 马尔科夫随机场, 遥感影像

中图分类号:

TP751.1

王心哲,邓棋文,王际潮,范剑超. 深度语义分割MRF模型的海洋筏式养殖信息提取[J]. 山东大学学报 (工学版), 2022, 52(2): 89-98.

WANG Xinzhe, DENG Qiwen, WANG Jichao, FAN Jianchao. Deep semantic segmentation MRF model for information extraction of marine floating raft aquaculture[J]. Journal of Shandong University(Engineering Science), 2022, 52(2): 89-98.

参考文献

[1] 王蒙蒙,李国庆,刘逸洁,等. 近20年来山东半岛东部海水养殖水面的动态变化[J]. 应用海洋学学报, 2017, 36(3): 319-326. WANG Mengmeng, LI Guoqing, LIU Yijie, et al. Dynamic changes of mariculture areas in eastern Shandong Peninsula in recent 20 years[J]. Journal of Applied Oceanography, 2017, 36(3): 319-326.
[2] 吕兑安,程杰,莫微,等. 海水养殖污染与生态修复对策[J]. 海洋开发与管理, 2019, 36(11): 43-48. LÜ Duian, CHENG Jie, MO Wei, et al. Pollution and ecological restoration of mariculture[J]. Ocean Develo-pment and Management, 2019, 36(11): 43-48.
[3] OTTINGER Marco, CLAUSS Kersten, KUENZER Claudia. Aquaculture: relevance, distribution, impacts and spatial assessments: a review[J]. Ocean & Coastal Management, 2016, 119(6): 244-266.
[4] 刘丛力,刘世禄. 我国海水养殖业发展现状与可持续发展问题[J]. 黄渤海海洋, 2001, 19(3): 100-105. LIU Congli, LIU Shilu. Maricultural development situ-ations and sustainable development problems in China[J]. Journal of Oceanography of Huanghai & Bohai Seas, 2001, 19(3): 100-105.
[5] 胡园园,范剑超,王钧. 广义统计区域合并的SAR图像浮筏养殖信息提取[J]. 中国图象图形学报, 2017, 22(5): 610-621. HU Yuanyuan, FAN Jianchao, WANG Jun. Modifying generalized statistical region merging for unsupervised extraction of floating raft aquaculture in SAR images[J]. Journal of Image and Graphics, 2017, 22(5): 610-621.
[6] 武易天. 基于遥感影像的近海岸水产提取方法研究[D]. 北京: 中国科学院大学(中国科学院遥感与数字地球研究所), 2017. WU Yitian. Research on coastal aquaculture detection using remote sensing images[D]. Beijing: University of the Chinese Academy of Sciences(Institute of Remote Sensing and Digital Earth, Chinese Academy of Sciences), 2017.
[7] 李俊杰,何隆华,戴锦芳,等. 基于遥感影像纹理信息的湖泊围网养殖区提取[J]. 湖泊科学, 2006, 18(4): 337-342. LI Junjie, HE Longhua, DAI Jinfang, et al. Extract enclosure culture in lakes based on remote sensing image texture information[J]. Journal of Lake Sciences, 2006, 18(4): 337-342.
[8] 周小成,汪小钦,向天梁. 基于ASTER影像的近海海水产养殖信息自动提取方法[J]. 湿地科学, 2006, 4(1): 64-68. ZHOU Xiaocheng, WANG Xiaoqin, XIANG Tianliang. Method of automatic extracting seaside aquaculture land based on ASTER remote sensing image[J]. Wetland Science, 2006, 4(1): 64-68.
[9] 徐珊,夏立华,彭海波,等. 基于面向对象的海水养殖模式遥感提取研究[J]. 测绘与空间地理信息, 2018, 41(5): 110-112. XU Shan, XIA Lihua, PENG Haibo, et al. Remote sensing extraction of mariculture models based on object[J]. Geomatics & Spatial Information Technology, 2018, 41(5): 110-112.
[10] 徐桃. 基于光谱数据空间结构特征分析的遥感蚀变信息提取研究[D]. 长沙: 中南大学, 2011. XU Tao. Remote sensing alteration information extraction based on spatial structure analysis of spectral data[D]. Changsha: Central South University, 2011.
[11] 徐雯佳. 基于高分卫星影像的秦皇岛近海浮筏养殖分布遥感监测[J]. 河北渔业, 2020(4): 32-34. XU Wenjia. Remote sensing monitoring of the distribution of offshore floating raft culture in Qinhuangdao based on high-score satellite imagery[J]. Hebei Fisheries, 2020(4): 32-34.
[12] 刘岳明,杨晓梅,王志华,等. 基于深度学习RCF模型的三都澳筏式养殖区提取研究[J]. 海洋学报, 2019, 41(4): 119-130. LIU Yueming, YANG Xiaomei, WANG Zhihua, et al. Extracting raft aquaculture areas in Sanduao from high-resolution remote sensing images using RCF[J]. Acta Oceanologica Sinica, 2019, 41(4): 119-130.
[13] 耿杰,范剑超,初佳兰,等. 基于深度协同稀疏编码网络的海洋筏式SAR图像目标识别[J]. 自动化学报, 2016, 42(4): 593-604. GENG Jie, FAN Jianchao, CHU Jialan, et al. Research on marine floating raft aquaculture SAR image target recognition based on deep collaborative sparse coding network[J]. Acta Automatica Sinica, 2016, 42(4): 593-604.
[14] 郑智腾,范海生,王洁,等. 改进型双支网络模型的遥感海水网箱养殖区智能提取方法[J]. 国土资源遥感, 2020, 32(4): 120-129. ZHENG Zhiteng, FAN Haisheng, WANG Jie, et al. An improved double-branch network method for intelligently extracting marine cage culture area[J]. Remote Sensing for Land & Resources, 2020, 32(4): 120-129.
[15] BADRINARAYANAN Vijay, KENDALL Alex, CIPOLLA Roberto. SegNet: a deep convolutional encoder-decoder architecture for image segmentation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(12): 2481-2495.
[16] LONG J, SHELHAMER E, DARRELL T. Fully convolutional networks for semantic segmentation[C] //Proceedings of IEEE Conference on Computer Vision and Pattern Recognition. Massachusetts, USA: CVPR Press, 2015: 3431-3440.
[17] SIMONYAN K, ZISSERMAN A. Very deep convolutional networks for large-scale image recognition[C] //Proceedings of International Conference on Learning Representations. California, USA: ICLR Press, 2014: 1-14.
[18] 郦苏丹,张翠,王正志. 基于马尔可夫随机场的SAR图象目标分割[J]. 中国图象图形学报, 2002, 7(8): 794-799. LI Sudan, ZHANG Cui, WANG Zhengzhi. SAR target segmentation based on Markov random field[J]. Journal of Image and Graphics, 2002, 7(8): 794-799.
[19] 柴震海,秦琴,王汝笠. 马尔可夫随机场在可见光图像分割中的应用[J]. 科学技术与工程, 2006, 6(6): 768-770. CHAI Zhenhai, QIN Qin, WANG Ruli. Markov's application in visible image segmentation with the airport[J]. Science Technology and Engineering, 2006, 6(6): 768-770.
[20] 曹兰英,夏良正,张昆辉. 基于小波域MRF模型的SAR图像分割[J].东南大学(自然科学版), 2004, 34(6): 847-850. CAO Lanying, XIA Liangzheng, ZHANG Kunhui. SAR image segmentation using MRF model in wavelet domain[J]. Journal of Southeast University(Natural Sciences Edition), 2004, 34(6): 847-850.
[21] 王蕊,王常颖,李劲华. 基于数据挖掘的GF-1遥感影像绿潮自适应阈值份去智能检测方法研究[J]. 海洋学报, 2019, 41(4): 131-144. WANG Rui, WANG Changying, LI Jinhua. Study of green tide adaptive thresholds for remote sensing images based on data mining[J]. Acta Oceanologica Sinica, 2019, 41(4): 131-144.
[22] 张涛,杨晓梅,童立强,等. 基于多尺度图像库的遥感影像分割参数优选方法[J] , 国土资源遥感, 2016, 28(4): 59-63. ZHANG Tao, YANG Xiaomei, TONG Liqiang, et al. Selection of best-fitting scale parameters in image segmentation based on multiscale segmentation image database[J]. Remote Sensing for Land & Resources, 2016, 28(4): 59-63.
[23] CHEN L C, PAPANDREOU G, KOKKINOS I, et al. DeepLab: semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected CRFs[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2018, 40(4): 834-848.

相关文章 15

[1]	张学思,张婷,刘兆英,江天鹏. 基于轻量型卷积神经网络的海面红外显著性目标检测方法[J]. 山东大学学报 (工学版), 2022, 52(2): 41-49.
[2]	蒋桐雨,陈帆,和红杰. 基于非对称U型金字塔重建的轻量级人脸超分辨率网络[J]. 山东大学学报 (工学版), 2022, 52(1): 1-8, 18.
[3]	吴建清,宋修广. 同步定位与建图技术发展综述[J]. 山东大学学报 (工学版), 2021, 51(5): 16-31.
[4]	柴庆发,孙守晶,邱吉福,陈明,魏振,丛伟. 气象灾害条件下电网应急物资预测方法[J]. 山东大学学报 (工学版), 2021, 51(3): 76-83.
[5]	陶亮,刘宝宁,梁玮. 基于CNN-LSTM 混合模型的心律失常自动检测[J]. 山东大学学报 (工学版), 2021, 51(3): 30-36.
[6]	杨修远,彭韬,杨亮,林鸿飞. 基于知识蒸馏的自适应多领域情感分析[J]. 山东大学学报 (工学版), 2021, 51(3): 15-21.
[7]	廖锦萍,莫毓昌,YAN Ke. 基于C-LSTM的短期用电预测模型和应用[J]. 山东大学学报 (工学版), 2021, 51(2): 90-97.
[8]	刘帅,王磊,丁旭涛. 基于Bi-LSTM的脑电情绪识别[J]. 山东大学学报 (工学版), 2020, 50(4): 35-39.
[9]	廖南星,周世斌,张国鹏,程德强. 基于类激活映射-注意力机制的图像描述方法[J]. 山东大学学报 (工学版), 2020, 50(4): 28-34.
[10]	蔡国永,贺歆灏,储阳阳. 基于空间注意力和卷积神经网络的视觉情感分析[J]. 山东大学学报 (工学版), 2020, 50(4): 8-13.
[11]	李春阳,李楠,冯涛,王朱贺,马靖凯. 基于深度学习的洗衣机异常音检测[J]. 山东大学学报 (工学版), 2020, 50(2): 108-117.
[12]	宋士奇,朴燕,蒋泽新. 基于改进YOLOv3的复杂场景车辆分类与跟踪[J]. 山东大学学报 (工学版), 2020, 50(2): 27-33.
[13]	蔡国永,林强,任凯琪. 基于域对抗网络和BERT的跨领域文本情感分析[J]. 山东大学学报 (工学版), 2020, 50(1): 1-7,20.
[14]	陈德蕾,王成,陈建伟,吴以茵. 基于门控循环单元与主动学习的协同过滤推荐算法[J]. 山东大学学报 (工学版), 2020, 50(1): 21-27,48.
[15]	张继,金翠,王洪元,陈首兵. 基于奇异值分解行人对齐网络的行人重识别[J]. 山东大学学报 (工学版), 2019, 49(5): 91-97.

多维度评价

Viewed

Full text

Abstract

Cited

Shared

Discussed