Information extraction of coastal aquaculture ponds based on spectral features and spatial convolution
LI Yefan1,2,3(), WANG Lin1,2,3(), ZHANG Dongzhu1,2,3
1. College of Environmental and Safety Engineering, Fuzhou University, Fuzhou 350108, China 2. Academy of Geography and Ecological Environment, Fuzhou University, Fuzhou 350108, China 3. Fujian Provincial Key Laboratory of Remote Sensing of Soil Erosion and Disaster Prevention, Fuzhou 350108, China
为控制水产养殖塘无序发展带来的负面效应,促进水产养殖业进一步发展,首要解决的就是对其快速、准确识别和提取的问题。水产养殖塘是被复杂道路和堤坝分割的特殊网状水体,单纯的光谱特征或空间纹理特征都不足以对其准确提取,且混合特征规则集对计算机性能要求越发苛刻。鉴于此,以Landsat影像序列为数据源,基于谷歌地球引擎(Google Earth Engine, GEE)平台,提出了一种结合影像光谱信息、空间特征和形态学操作的沿海水产养殖塘自动提取方法。该方法联用了双特征水体光谱指数(改进型组合水体指数(modified combined index for water identification,MCIWI)与改进的归一化差异水体指数(modified normalized difference water index,MNDWI))以突出大面积水体与养殖塘的网格特征,再利用低频滤波空间卷积运算拉伸养殖与非养殖水体之间的差异特征,将水产养殖塘区作为一个整体准确识别和快速提取。研究结果表明: ①该方法总精度达到93%,Kappa系数为0.86,典型区域叠加比对检验流程验证,提取结果和实际结果重叠比例均在90%以上,平均重叠比例达92.5%,反映了提取方法的高精度和可靠性; ②2020年福建省近岸海域水产养殖塘区总面积为511.73 km2,主要分布在漳州市、福州市和宁德市; ③核密度分析结果表明漳州市的水产养殖塘集聚度高,相应其养殖塘管理压力也较大。该方法可以实现近岸海域水产养殖塘的自动化提取,对促进渔业养殖的有序管理和科学发展具有重要的意义。
To control the negative effects of the disorderly development of aquaculture ponds and promote the further development of the aquaculture industry, the top priority is to realize rapid and accurate identification and extraction of information on aquaculture ponds. Aquaculture ponds are special net-like water bodies divided by complex roads and dikes. Simple spectral features or spatial texture features are not enough for accurate information extraction. Moreover, the mixed feature rule set gets more demanding on computer performance. Therefore, based on the Landsat image sequence and the Google Earth Engine (GEE) platform, this study proposed an automatic extraction method for coastal aquaculture ponds, which combined the image spectral information, spatial features, and morphological operation. In this method, dual characteristic water spectral indices, that is, the modified combined index for water identification (MCIWI) and the modified normalized difference water index (MNDWI), were employed to highlight the grid characteristics of large water bodies and aquaculture ponds. Then, the low-frequency filtering spatial convolution operation was used to stretch the differences between aquaculture and non-aquaculture water bodies. Finally, the information on aquaculture pond areas as a whole was identified and extracted accurately and quickly. The results are as follows. ① This method has overall precision of 93% and a Kappa coefficient of 0.86. According to the test process verification of typical area superposition comparison, the overlapping proportions between the extraction results and the actual results were all more than 90%, averaging 92.5%, reflecting the high precision and reliability of this extraction method. ② In 2020, the coastal aquaculture ponds in Fujian Province had a total area of 511.73 km2and were mainly distributed in Zhangzhou, Fuzhou, and Ningde cities. ③ The kernel density analysis suggested that Zhangzhou had a high concentration of aquaculture ponds and thus had high pressure in the management of aquaculture ponds. This method can realize automatic information extraction of coastal aquaculture ponds. Thus, it is of great significance to promote the orderly management and scientific development of fishery aquaculture.
李叶繁, 王琳, 张冬珠. 光谱特征和空间卷积相协同的近岸海域养殖塘遥感信息提取[J]. 自然资源遥感, 2022, 34(4): 42-52.
LI Yefan, WANG Lin, ZHANG Dongzhu. Information extraction of coastal aquaculture ponds based on spectral features and spatial convolution. Remote Sensing for Natural Resources, 2022, 34(4): 42-52.
Wang M M, Li G Q, Liu Y J, et al. Dynamic changes of mariculture surface in eastern Shandong Peninsula during the past 20 years[J]. Journal of Applied Oceanography, 2017, 36(3):319-326.
Shi Y, Su T, Lu L R. Remote sensing monitoring of temporal and spatial evolution of seine farming area in Changdang Lake[J]. Environmental Monitoring and Warning, 2021, 13(3):8-12.
Cao L, Gu W J, Li X S, et al. Remote sensing investigation of oyster culture distribution in Tiegang,Zhejiang Province based on WorldView data[J]. Fisheries Information and Strategy, 2016, 31(4):286-292.
[8]
Chen L, Ren C, Zhang B, et al. Spatiotemporal dynamics of coastal wetlands and reclamation in the Yangtze estuary during past 50 years(1960s—2015)[J]. Chinese Geographical Science, 2018, 28(3):386-399.
doi: 10.1007/s11769-017-0925-3
[9]
Zhou M, Wu M, Zhang G, et al. Analysis of coastal zone data of northern Yantai collected by remote sensing from 1990 to 2018[J]. Applied Sciences, 2019, 9(20):4466.
doi: 10.3390/app9204466
[10]
Xu M, Cui B, Lan S, et al. Exploring dynamic change of the tidal flat aquaculture area in the Shandong Peninsula (China) using multitemporal Landsat imagery(1990—2015)[J]. Journal of Coastal Research, 2020, 99:197-202.
doi: 10.2112/SI99-028.1
Shao Y T, Lu X, Ye H, et al. Object-oriented classification and monitoring of land cover change in coastal wetland[J]. Marine Science, 2020, 44(1):52-66.
Meng J, Hua C, Zhang J B, et al. Spatial and temporal distribution of aquaculture land in Guangxi coastal zone from 1995 to 2015[J]. Guangxi Science, 2017, 24(2):135-143.
Li Y, Zhang C F, Shuai S, et al. Research on extraction method of offshore aquaculture information in Weihai City[J]. Wetland Science, 2017, 15(5):651-656.
[14]
Sun N, Zhu W, Cheng Q. GF-1 and Landsat observed a 40-year wetland spatiotemporal variation and its coupled environmental factors in Yangtze River estuary[J]. Estuarine,Coastal and Shelf Science, 2018, 207:30-39.
doi: 10.1016/j.ecss.2018.03.022
Wang N, Cheng J H, Zhang H Y, et al. Remote sensing dynamic monitoring of aquaculture water and its application[J]. Journal of Fishery Sciences of China, 2019, 26(5):893-903.
Cheng B, Liu Y M, Liu X N, et al. Research on extraction method of offshore aquaculture area based on multi-source feature fusion of high-resolution remote sensing image[J]. Remote Sensing Technology and Application, 2018, 33(2):296-304.
Huang S, Song K H, Luo J H, et al. Remote sensing extraction algorithm for shallow lake seine area based on gradient transformation[J]. Journal of Lake Sciences, 2017, 29(2):490-497.
doi: 10.18307/2017.0225
Sun Y Y, Ying J L, Wang M, et al. Study on classification extraction and dynamic monitoring of marine information for aquaculture[J]. Marine Information, 2016(2):6-11.
Huang S W, Wei C Z. Changes of artificial aquaculture bases in coastal cities of Guangdong Province:An empirical study based on Sentinel-1 data from 2015 to 2019[J]. Tropical Geography, 2021, 41(3):622-634.
Pei L, Wang J X, Qu H H, et al. Remote sensing image extraction of coastal aquaculture ponds based on ONDPI[J]. Marine Surveying and Mapping, 2020, 40(5):40-44.
Yuan X, Zhang L, Song Q Q, et al. Remote sensing monitoring and spatial-temporal analysis of seawater pond aquaculture in Hainan Island[J]. Journal of Oceanographic Research, 2020, 38(1):59-67.
Wang F, Xia L H, Chen Z B, et al. Object-oriented mariculture pattern recognition based on association rules[J]. Transactions of the Chinese Society of Agricultural Engineering, 2018, 34(12):210-217.
[23]
Fu Y, Deng J, Ye Z, et al. Coastal aquaculture mapping from very high spatial resolution imagery by combining object-based neighbor features[J]. Sustainability, 2019, 11(3):637.
doi: 10.3390/su11030637
[24]
Gorelick N, Hancher M, Dixon M, et al. Google Earth Engine:Planetary-scale geospatial analysis for everyone[J]. Remote Sensing of Environment, 2017, 202:18-27.
doi: 10.1016/j.rse.2017.06.031
Yang P, Tan L S, Huang J F, et al. Diurnal variation of CH_4 and N_2O fluxes after drainage from aquaculture ponds in the Minjiang River estuary during early winter[J]. Environmental Science, 2018, 39(1):300-309.
[26]
Ren C, Wang Z, Zhang Y, et al. Rapid expansion of coastal aquaculture ponds in China from Landsat observations during 1984—2016[J]. International Journal of Applied Earth Observation Geoinformation, 2019, 82:101902.
doi: 10.1016/j.jag.2019.101902
[27]
Duan Y, Li X, Zhang L, et al. Detecting spatiotemporal changes of large-scale aquaculture ponds regions over 1988—2018 in Jiangsu Province,China using Google Earth Engine[J]. Ocean and Coastal Management, 2020, 188:105144.
doi: 10.1016/j.ocecoaman.2020.105144
[28]
Duan Y, Li X, Zhang L, et al. Mapping national-scale aquaculture ponds based on the Google Earth Engine in the Chinese coastal zone[J]. Aquaculture, 2020, 520:734666.
doi: 10.1016/j.aquaculture.2019.734666
Yang B G, Chen F, Luo Z Z. Extraction of complex water information based on MODIS improved combined water index (MCIWI)[J]. Journal of Southwest University (Science and Technology Edition), 2011, 33(1):112-119.
Chen D L, Liu J N, Yu L L. Comparison and research on several methods of image segmentation threshold selection[J]. Machine Building and Automation, 2003(1):77-80.
Li B Y, Fan Y G, Gao Y. Infrared image feature extraction based on OTSU and Canny operator[J]. Journal of Shaanxi University of Technology (Natural Science Edition), 2019, 35(6):33-40.
[32]
Zhang G, Wu M, Wei J, et al. Adaptive threshold model in Google Earth Engine:A case study of ulva prolifera extraction in the South Yellow Sea,China[J]. Remote Sensing, 2021, 13(16):3240.
doi: 10.3390/rs13163240
[33]
Donchyts G, Schellekens J, Winsemius H, et al. A 30 m resolution surface water mask including estimation of positional and thematic differences using Landsat8,SRTM and openstreetmap:A case study in the Murray-Darling basin,Australia[J]. Remote Sensing, 2016, 8(5):386.
doi: 10.3390/rs8050386
Xu H Q. Study on water information extraction using improved normalized difference water index (MNDWI)[J]. Journal of Remote Sensing, 2005, 9(5):589-595.