小麦种植与生产对于保障国家粮食安全和维护社会稳定具有重要意义.苗情和施肥管理对冬小麦的生长发育、产量和品质具有重要作用.针对传统苗情调查方法存在的成本较高、时效性差等问题,提出了一种基于覆盖度特征和纹理特征的冬小麦越冬期苗情等级分类方法,该方法通过图像就能够快速获取对应的苗情等级.通过将采样点的苗情等级与遥感影像的归一化差值植被指数NDVI相结合,实现了"点状"苗情等级向"面状"苗情等级的映射,得到研究区的冬小麦苗情分级结果.根据苗情分级结果和冬小麦的需肥规律,制定出对应的变量施肥决策.试验结果表明,所提出的基于覆盖度特征和纹理特征的冬小麦越冬期苗情等级分类方法的识别准确率达到90%,并且基于两种特征比基于单一特征的冬小麦越冬期苗情等级分类方法的识别准确率高.实地验证表明,基于采样图像和遥感影像得到的研究区冬小麦苗情分级结果准确率达到92%.相较于传统的施肥方法,采用提出的变量施肥决策对研究区冬小麦进行施肥时,平均每公顷减少化肥用量209.30 kg.
Seedling Condition Grading and Variable Fertilization Method of Winter Wheat Based on Multi-source Data
Wheat planting plays a significant role in ensuring national food security and maintaining social stability.The seedling condition and fertilization management of winter wheat is crucial to its growth,development,yield,and quality.Aiming at the problems of high cost and poor timeliness of traditional seedling condition survey methods,a classification method for winter wheat seedling condition grading during the overwintering period based on coverage and texture features is proposed.This method allows for the rapid acquisition of corresponding seedling condition grades through images.By combining the seedling condition grades of sampling points with the normalized difference vegetation index(NDVI)from remote sensing images,the mapping of"point"seedling condition grades to"area"seedling condition grades has been achieved,and the winter wheat seedling condition grading results for the study area are obtained.Based on the seedling condition grading results and the fertilization patterns of winter wheat,the corresponding variable fertilization decisions are formulated.Experimental results show that the identification accuracy of the proposed winter wheat seedling condition classification method during the overwintering period based on coverage and texture features reaches 90%,and the identification accuracy of the method based on two features is higher than that based on a single feature.Field verification shows that the accuracy of the winter wheat seedling condition grading results for the study area obtained from sampling images and remote sensing images reaches 92%.Compared with the traditional fertilization method,when the proposed variable fertilization decisions are adopted to fertilize winter wheat in the study area,the average reduction of chemical fertilizer usage by per hectare is 209.30 kg.
NETL
NSTL
万方数据
