基于机器视觉的马体尺测量系统设计与研究

Design and Research of Measurement System of Horse Body Based on Machine Vision

摘要:

传统马体尺的人工测量方法工作量大且存在安全隐患,对此提出基于线性回归理论和机器视觉技术的马体尺测量方法,旨在测量马体的基本数据如体高、体长、胸围、管围;首先,在Matlab中利用图像腐蚀方法得到马体轮廓,并在2D图像上精确定位马体坐标,获得体高、体长指标；然后,自定义胸径、管径指标,代入线性回归方程预测胸围、管围；最后利用Matlab GUI工具设计系统可视化界面,并初步完成系统的仿真测试;仿真结果表明,利用线性相关及线性回归理论解决3D指标的预测问题,具备测量依据和借鉴意义。

Abstract:

The traditional manual measuring method of horse body usually have large workload and security risks.Based on linear regression theory and machine vision technology,we put forward the new measuring method of horse body. The basic data is designed to measure the horse such as body height, body length, chest length, vessel length.Firstly, the image is used to get the horse body contour by image erosion based on Matlab, and the body coordinate is accurately located on the 2D image, so we can obtain the body height and body length index. then, diameter of the chest and diameter of the vessel were defined to predict chest length and vessel length by plugging into equation of linear regression; Finally,we use the Matlab GUI to design the visual interface of the system, and complete the simulation test of the system. The simulation results show that the linear correlation and linear regression theory which can be used to solve the prediction problem of 3D index has the basis of measurement and reference.