基于机器视觉的种子间距识别及其应用

摘要

介绍了基于机器视觉的精播排种器性能检测方法,在对获取的样本图像进行处理的基础上完成帧种子间距特征的识别.通过分析种子运动特征,实现了排种时间间隔样本的提取.利用理论排种和实际排种时间间隔之间的关系,构建了以时间间隔为特征量的排种质量统计模型,计算出排种器性能指标.试验表明,此方法检测误差小、速度快、精度高.%The machine vision method was adopted to detect the performance of precision seedmeter. Through a series of treatments, the frames of seed spacing were recognized. The samples of the seed interval time were extracted by analyzing of seed movement. The statistical model was deduced based on the relationship of the theory interval time and the practical interval time. Finally, the performance of the seedmeter was counted. The result of experiment showed that the method precision and velocity were satisfactory.