论文部分内容阅读
用ＮＯＡＡ卫星资料对１９９８年吉林省西部地区洪涝灾害进行了动态监测，并以农田损失为主对灾情进行了评估。采用ＡＶＨＲＲ通道１与通道２的比值图像突出水体信息，在直方图上人机交互划定水陆边界，用统计数学方法确定了混合像元中水体与非水体的阈值。利用灾区当年的作物种植数据作出农田受灾面积评估，用受灾前后的植被指数图，进行农田受灾程度划分。对像元面积的精确计算可以提高受灾面积的计算精度，区域像无数的统计分析可以减少区域内灾害面积比例分析的误差。
Using the NOAA satellite data to monitor the flood disaster in the western region of Jilin Province in 1998, the disasters were evaluated based on farmland loss. The ratio of AVHRR channel 1 to channel 2 is used to emphasize the water information, and the human-computer interaction is used to demarcate the water-land boundary in the histogram. The thresholds of water bodies and non-water bodies in mixed pixels are determined by statistical methods. Make use of the crop planting data of the affected area to make an assessment of the affected area of farmland and make use of the vegetation index before and after the disaster so as to classify the degree of farmland disaster. Accurately calculating the pixel area can improve the calculation accuracy of the affected area. Countless statistical analysis of the area can reduce the error of proportional analysis of the disaster area in the area.