论文部分内容阅读
为了提高棉花苗情监测的时效性和精确性,本研究利用无人机低空获取棉田数字图像,通过图像分析快速识别诊断棉花苗情。研究结果表明,利用HIS(Hue-intensity-saturation)阈值法将图像二值化,之后通过腐蚀膨胀对二值化图像进行处理,能够较好地排除地膜干扰,快速识别大范围棉苗数量和壮苗数量,棉苗识别精度超过90%。基于图像识别结果绘制的田间苗情分布图清晰显示了棉田出苗情况,并为精准化管理提供依据。本研究结果可为无人机在农业中的应用提供参考。
In order to improve the timeliness and accuracy of cotton seedling monitoring, this study uses the low altitude of drone to obtain the digital images of cotton fields, and quickly diagnoses the cotton seedlings by image analysis. The results show that using HIS (Hue-intensity-saturation) thresholding method to binarize the image, and then processing the binarized image through erosion and swelling can better eliminate the mulching and quickly identify the number and strength Seedling number, cotton seedling recognition accuracy of more than 90%. Based on the results of image recognition, the plot of seedling distribution in the field clearly shows the seedling emergence in cotton field and provides the basis for accurate management. The results of this study may provide a reference for the application of UAV in agriculture.