粘胶长丝在线实时监测系统设计与处理算法

介绍了一种可在酸浴内实时监测粘胶长丝数目的检测系统及图像处理算法。系统采用柱面LED光源照明,经CCD摄像头采集图像至计算机并进行图像处理,实时计算出单丝的数量,给出是否合格的判别结果。采用基于人眼视觉特性的自适应图像增强处理算法,对粘胶长丝图像进行增强,利用类间最大交叉熵的阈值分割法和连通域跟踪计数法,解决了粘胶长丝计数中的胶丝重叠、分叉等关键技术难题,保证了计数准确率。实验证明,该系统具有精度高、检测速度快等优点,适合在化纤企业推广,具有良好的应用前景。

Viscose filament real time detection system design and image processing algorithm

This essay introduces a viscose filament detection system which can conduct the real time monitoring in acid solution slot on the quantity of viscose filament ejected,and correlated image processing algorithm.This system applies cylinder LED lighting source,via CCD camera is entered into a computer.Computer can calculate the quantity of single filament via image processing algorithm,and identify its quality whether qualified or not.Regarding image processing,viscose filament image is enhanced via image enhancement algorithm,the enhancement is adaptive image enhancement algorithm based on human visual properties.By applying inter-class cross entropy threshold segmentation method and connected domain tracking algorithms,it solves the technical problems of the overlapping and bifurcation in viscose filament counting and increases the accuracy.Experiments prove that this system has the advantages of high accuracy and fast detection.It can be applied broadly in chemical fabrics industry.