基于激光雷达的猪只体重测量系统开发与研究
2022,30(3):88-94
摘要：随着养殖业规模化的不断发展，自动化养猪已成为目前猪只养殖业的一种趋势。针对传统猪只体重测量过程中出现一系列应激性行为问题，本文开发研究了一款基于激光雷达猪只体重测量系统，该系统通过以激光雷达为核心的硬件设备与软件系统相结合，实现了猪只二维模型建立；在此基础上采用QT3D框架构建视窗，并对视窗进行水平垂直布局并构建猪只三维模型视窗。采用Delaunay三角划分的方式叠加得到猪只体积，利用系统获得体积数据与猪只真实体重构建体积与体重数学模型，拟合精度R2为0.92535。对该模型验证分析得知体重误差在±8%范围内。该系统扫描速度快、建模效果良好、体重计算精确，为猪只非接触式测重和相关领域的点云数据三维建模应用提供了新的技术研究方法。
关键词：猪只;激光扫描;数据预处理;三维建模;体重计算
Development and Research of Pig Weight Measurement System Based on Laser Thunder Based on Laser Radar
Abstract：With the continuous development of large-scale breeding industry, automated pig breeding has become a trend of pig breeding industry. In view of a series of stress behavior problems in the process of traditional pig weight measurement, this paper developed and studied a a Lidar Measurement System for Body Weight. Through the combination of hardware and software system with lidar as the core, the two-dimensional pig model was established. On this basis, The QT3D frame is used to construct the windows, and the windows are laid out horizontally and vertically to construct the 3D model windows of pigs. The volume of pigs was superimposed by Delaunay trigonometry. Using the volume data collected by the system to construct a mathematical model of volume and weight, the fitting accuracy R2 is 0.92535. Validation analysis of the model shows that the weight error is within ±8%. The system has the advantages of fast scanning speed, good modeling effective and accurate weight calculation. It provides a new technical research method for pig non-contact weight measurement and 3D modeling of point cloud data in related fields.
Key words：the pigs; laser scanning; data preprocessing; three-dimensional modeling; weight calculation
收稿日期：2022-01-27
基金项目：国家自然科学(31772651)；国家重点研发计划项目(2018YFD050070503)。
