摘要
介绍了一种基于惯性主轴的水果分级目标姿态检测方法。采用计算惯性主轴的方法求取水果的果轴信息,在测试台上通过增加一个反光镜面的方法获取同一水果两个视角的图像,分别求出各图像的惯性主轴,据此获取空间水果目标的姿态。结果表明,正投影果轴倾角实测值与计算值的平均误差为4.4°,果轴检测正确率为86%,果轴与测试台夹角实测值与计算值的平均误差为5.9°,该方法可有效检测分级生产中水果的姿态。
A detection method for object attitude based on principal axis of inertia which was applied in fruit grading was introduced.Fruit axis information was calculated by calculating the axis of inertia and two perspective images of the same fruit were obtained by adding a reflector,and the principal axis of inertia of each image was obtained respectively to obtain the attitude information of the space fruits object.The experiment results showed that the average error of the orthographic projection fruit axis angle between measured value and calculated value was 4.4°,the correct rate of fruit axis detection was 86%,and the average error of the angles of fruit axis and test bench between measured value and calculated value was 5.9°.This method was favorable to detect the object attitude of fruit grading.
出处
《湖北农业科学》
北大核心
2012年第17期3862-3865,共4页
Hubei Agricultural Sciences
基金
浙江省教育厅科研项目(Y201017164)
关键词
果轴
惯性主轴
目标姿态
fruit axis
principal axis of inertia
object attitude
