基于YOLOv10的烤烟种植面积快速提取方法

打开文本图片集

中图分类号：S572 文献标志码：A 文章编号：2096-9902（2026）04-0025-04

Abstract：Inordertoquicklyextract theplanting rangeofflue-cured tobacco，thispaper proposesa method basedon UAV aerialphotography.TheYOLOv1Otargetdetectionalgorithmisusedtoquicklyscreenthephotoscontainingtheflue-cured tobaccoplantingareafromtheaerialphotography，andthephotosaresplicedtoproduceorthophotoimagesoftheflue-cured tobaccoplantingarea.quicklyextracttheplantingrangeofflue-curedtobacothroughimageinterpretation.Theexperimentused theYOLOv10algorithmtoselect577photoscontainingflue-curedtobacoplantingareasfrom1889UAVaerialphotosto produceorthophotoimagesofflue-curedtobaccoplantingareas，whichreducedthetimerequired toproduceorthophotoimages andrealizedrapidpositioningoftheflue-curedtobaccoplantingrangeandeliminatedalargeamountofinterferenceinformation fromvisualinterpretationofimageinterpretationExperimentalresults：Theorthophotoimagesoftheflue-curedtobaccoplanting areaproducedbythismethodwerecombinedwiththemeasuredflue-curedtobaccoplantingrange.Theerorareawas O.6mu and the relative error was 0.3% . The rapid extraction method of flue-cured tobacco planting range proposed in this paper can quickly and eficiently determine the flue-cured tobacco planting range.

Keywords：planting range; YOLOv1O;flue-cured tobacco identification;area extraction;aerial photography

烤烟作为云南省重要的农业收入来源，2023年云南省烤烟种植面积达到623万亩（1亩约等于 667m2 ，下同），烟草产业实现工商业合计利税1834万元，全省烟农总收人预计达到260亿元1。（剩余3464字）