基于多特征的Gaofen-1海冰影像监督分割

Study on Supervised Segmentation Based on Multi-featured Gaofen-1 Sea Ice Image

冬季沿海地区的海冰检测工作对该地区居民的生产生活具有重要的指导意义,同时可以根据海冰变化检测气候变暖情况。海冰影像分割是海冰检测的基础。在众多海冰影像数据源中,Gaofen-1海冰影像因其丰富的光谱特征、较高的空间分辨率、简单的数据结构,在变化监测中具有重要应用价值。本文提出了一种基于红绿蓝3个光谱通道的灰度共生矩阵,提取遥感影像的纹理特征和光谱特征构成多特征的Gaofen-1海冰影像监督分割方法。以Gaofen-1合成的模拟海冰影像和某海湾地区真实Gaofen-1海冰影像进行分割实验,实验结果很好地证明了算法的可行性和可靠性。

Sea ice monitoring in coastal areas in winter has great importance in guiding the life and production.It is also of great significance in monitoring global warming according to the sea ice changes.Among the sea ice data sources,the Gaofen-1 image has important application values on monitoring changes with its advantages of rich spectral characteristics,high spatial resolution,and simple data structure.A Gray-level Co-occurrence Matrix(GLCM)based on three spectral channels(red,green,and blue)is proposed in this paper.The textural and spectral features of the Gaofen-1 sea ice image are extracted to form multi-feature to realize the supervised segmentation of images.Simulated sea ice images synthesized by Gaofen-1 remote sensing image,and the real Gaofen-1 sea ice images in the Liaodong bay are segmented by the proposed method.The experimental results prove that the proposed method is feasible and reliable in the segmentation of the sea ice image.