中文摘要：

梳棉机棉结检测系统是基于机器视觉理论提出的一种新型在线检测系统。光源在系统中起到照明的作用，是影响棉结检出率的重要因素。为满足棉结在线检测系统宽幅面、高速和高分辨率的检测要求，需要对光源系统进行优化设计。通过实验可以获知棉网上的棉结对何种波长范围的光具有最大的反射率，以及棉结和周边结构如何获得最好的对比度。在比较了传统的荧光灯、高压汞灯和发光二极管（LED）线光源的光谱曲线后，提出了采用He-Ne激光器与光学系统配合后进行照明的光源设计方案。实测结果表明，由于激光束的高亮度和高准直性的优点，这种光源较其他光源更适合棉结检测照明的实际环境，可以满足检测需要。

英文摘要：

Nep detection system used in carding machine is a kind of new on line detection system based on machine vision. The function of light source is illumination. Light source is an important factor for cleaning rate. The light source system should be optimized and designed to satisfy the broad surface, high speed and high sensitivity for online nep detection system. By experiments, we could know that the neps in cotton web would have the highest reflectivity illuminated by which kind of light source. Synchronously, we could know that the neps and surrounding structure how to get the best contrast. Compared with the spectral curve of traditional fluorescence lamp, high pressure mercury lamp and LED, the scheme of combination in He-Ne laser and optical system was adopted in this paper. Experimental results show that this light source, for the circumstance of nep detection in carding machine, is better than other light sources because of the high brightness and straightness of laser.