基于多光谱遥感的耕地等别识别评价因素研究

为了判断应用遥感技术进行耕地质量识别的可行性,该文以多光谱遥感影像为数据源,采用遥感技术提取耕地质量影响因素,进行耕地质量遥感识别评价。通过建立反演模型和实地验证,有机质含量、地形坡度与表层土壤质地可以利用遥感影像有效获取;灌溉和排水因素理论上可以通过遥感影像识别,但资料获取较困难,可通过专题图和影像解译相结合的方法获取;剖面构型、障碍层次和土壤酸碱度等因素通过遥感识别的准确度较低。以自然等指数作为遥感识别评价成果,利用已有农用地分等成果对其进行验证,结果表明,在误差范围7%,7%]内二者具有高度一致性,因此基于多光谱遥感进行耕地质量识别评价具有一定的可行性。该研究可为耕地等别评价因素识别、耕地质量评价和动态监测提供参考。

Evaluation factors for cultivated land grade identification based on multi-spectral remote sensing

To determine the feasibility of using remote sensing technique to identify cultivated land quality,multi-spectral remote sensing image was taken as data source and remote sensing technique was used to extract factors、evaluate cultivated land quality in this paper.Soil organic matter、terrain slope and surface texture were effectively obtained from remote sensing images through constructing inversion model and on-the-spot investigation;In theory,irrigation and drainage factors could identify through remote sensing images,but it was more difficult to obtain them,so they could be got by methods of Thematic map and Image interpretation;the factors obtained by identification of remote sensing,such as Profile configuration、obstacles and soil pH,had lower accuracy.Natural index was taken as evaluation results of RS identification and Agricultural Land Classification results existed was used to verify it,the result showed that they had a high degree of consistency within an error range between plus seven to minus seven.So it had certainly feasibility to identify cultivated land quality assessment by using multi-spectral remote sensing.The study could provide some references for identification of cultivated land grade evaluation factors、cultivated land quality evaluation and dynamic monitoring.