一种利用TM影像自动提取细小水体的方法

A Method of Small Water Information Automatic Extraction from TM Remote Sensing Images

提出一种利用TM影像自动提取山区细小水体的多波段谱间关系改进方法。该方法在典型谱间关系法的基础上,针对水体与阴影在蓝绿光波段亮度值降低速率差异较大的特征,基于差值运算,构建新的多波段谱间关系水体提取模型。首先利用该模型将水体从其他地物及阴影中分离出来,然后基于数学形态学膨胀滤波算法进行空洞填充和短线连接,最后通过图像细化算法实现目标的细化。经过试验比较表明,该方法克服了许多水体提取模型只能有效提取较大面积水体的缺点,除了能够对山区的细小水体进行高精度自动提取外,还能够有效地去除阴影等干扰信息。

Since the decrease rate of bright values of water body and shadow in blue-green bands are significantly different,the proposed approach build a novel water body extraction model of multi-band spectral relationship based on difference operation.Firstly,the approach utilizes the new model to extract water body among various ground objects and shadows.Secondly,the approach fills empty holes and connects short lines based on dilation filtering algorithm of mathematical morphology.Thirdly,the approach utilizes binary image thinning algorithm to thin the extracted water bodies.The analysis and comparison of experimental results show that the proposed approach overcomes the major shortcoming of many water body extraction approaches,which are only able to extract large water bodies.In summary,the approach is able to automatically extract small water bodies of hilly area with high precision.Moreover,it is able to effectively remove interference information such as shadows.