基于U 2-Net深度学习模型的沿海水产养殖塘遥感信息提取

A method for extracting information on coastal aquacultural ponds from remote sensing images based on a U 2-Net deep learning model

针对近海沿岸复杂地理环境中“同谱异物”效应导致传统方法提取水产养殖塘边界模糊、精度较低的问题,提出了基于U 2-Net深度学习模型的沿海水产养殖塘遥感信息提取方法。首先,对遥感影像进行预处理,选择合适的波段组合方式以区分养殖塘和其他地物;其次,通过目视解译进行样本制作;然后,利用U 2-Net深度学习模型训练并提取沿岸养殖塘;最后,利用局部最佳法确定养殖塘范围。实验结果表明,该方法平均总体精度达到95.50%,平均Kappa系数、召回率和F值分别为0.91,91.45%和91.01%;在养殖塘个数及面积评价方面,提取出养殖塘区19块,共计9.79 km^(2),区块数和面积的平均准确度分别为94.06%和93.18%。本研究能够快速、准确地开展海岸带区域养殖塘制图,能够为海洋资源管理和可持续发展提供技术支持。

Conventional information extraction methods for aquacultural ponds frequently yield blurred boundaries and low accuracy due to the effect of different objects with the same spectrum in complex geographical environments of offshore and coastal areas.This study proposed a method for extracting information on coastal aquacultural ponds from remote sensing images based on the U 2-Net deep learning model.First,an appropriate band combination method was selected to distinguish aquacultural ponds from other surface features through preprocessing of remote sensing images.Samples were then prepared through visual interpretation.Subsequently,the U 2-Net model was trained,and information on coastal aquacultural ponds extracted.Finally,the scopes of aquacultural ponds were determined using the local optimum method.The experimental results show that the method proposed in this study yielded the average overall accuracy of 95.50%,with the average Kappa coefficient,recall,and F-value of 0.91,91.45%,and 91.01%,respectively.Furthermore,19 ponds were extracted,with a total area of 9.79 km^(2).The average accuracies of the number and area of aquacultural ponds were 94.06%and 93.18%,respectively.The method proposed in this study allows for quick and accurate mapping of coastal aquacultural ponds,thus providing technical support for marine resource management and sustainable development.