Classification methods fitting for winter wheat planting area extraction in Hengshui area were studied based on HJ-1A remote sensing data. winter wheat planting areas in 2010, 2012 and 2010 in Hengshui were extracted using the maximum likelihood method, the minimum distance method, Mahalanobis distance method and ISODATA algorithm respectively, and were compared with real planting areas. The results showed that the average errors of winter wheat planting area extracted by maximum likelihood method, the Mahalanobis distance method, the minimum distance method and the ISODATA algorithm were 6.1%, 25.9%, 24.9% and 17.2%respectively. Maximum likelihood method presented higher estimation accuracy and can meet the requirements in studying the winter wheat planting area estimation.%采用HJ-1A遥感数据,研究了衡水市冬小麦种植面积提取的分类方法。分别运用最大似然法、最小距离法、马氏距离法和ISODATA算法对2010年、2012年和2013年衡水市冬小麦种植面积进行提取,并与实际种植面积进行比较分析。结果表明,最大似然法、马氏距离法、最小距离法和ISODATA法提取的冬小麦种植面积的平均误差分别为6.1%、25.9%、24.9%和17.2%。因此,最大似然法具有较高的估算精度,能够满足研究区域冬小麦种植面积估算要求。
展开▼