棉花颜色视觉检测系统及影响因素分析研究

Study on Cotton Color Vision Detection System and its Influencing Factors

针对在采用机器视觉技术测量棉花颜色等级时的外界光照影响严重、准确率较低的问题,提出了基于机器视觉的棉花颜色检测系统及其结构。设计视觉检测平台,避免外界光源噪声的影响。选取89C51单片机作为控制部分的核心,利用电机控制棉花的表面处理。软件部分通过采集棉花图像的三通道颜色,并利用颜色空间转化技术获取棉花颜色值,实现对棉花颜色值的在线检测。利用改变外界环境光及对棉花上表面的压力作为实验变化量,分别检测不同场景的棉花颜色值并加以分析。实验结果表明:在非密闭环境会对棉花颜色检测造成影响,且需要对棉花上表面施加至少40 N的压力,才能降低棉花因蓬松而对实验结果产生影响;在设计的装置下测量棉花颜色值的反射率误差率最低为0.06%、黄度误差率最低为0.35%。

Aiming at the problem of low accuracy caused by external illumination when measuring cotton color grade by machine vision technology,a cotton color detection system based on machine vision and its structure are presented.Visual inspection platform is designed to avoid noise from external light sources.Selecting 89C51 microcontroller as the core of the control part,the surface treatment of cotton is controlled by motor.The software collects the three channels color of cotton images,the cotton color value is obtained by color space transformation technology,realize the online detection of cotton color value is realized.Changing ambient light and pressuring on the upper surface of cotton is used as experimental variation,the cotton color values of different scenes are detected and analyzed.It is shown that non closed environment will affect cotton color detection,a pressure of at least 40 N on the upper surface of cotton is required to reduce the influence of cotton fluffiness on experimental results.Under the detection of the design device,the error rate of reflectivity of cotton color value can reach 0.06%and the error rate of yellowness can reach 0.35%.