基于机器视觉的尺寸检测智能工位设计

Design of Intelligent Workstation for Dimension Measurement Based on Machine Vision

摘要:

近年来,随着食品行业对质量要求的提升,食品加工行业逐步开始进行智能化转型,开始应用新型技术来解决生产过程中的检测问题。为了检测产线上不规则排布的冰棍的长度,设计了一种新型智能检测工位,并开发了配套算法。在算法开发中,融合了边缘检测及最小外廓拟合算法实现了连续非接触式的产品尺寸测量技术,并在7组238份样例的验证中,测量结果低于1.1%的标准相对误差,达到了理想的目标。

Abstract:

In recent years, as the requirements of hygiene and safety regulations have increased, the food industry has gradually begun to undergo intelligent transformation. Taking popsicles as an example, this paper designs a new type of intelligent detection station and matching algorithms. Using edge detection and minimum contour fitting, a non-contact measurement technology for product size was realized, and a standard relative error of less than 1.1% was obtained in 7 groups of 238 data sets.