基于机器视觉的输送带图像在线采集技术的研究

Research on On-Line Image Acquisition Techniques of Conveyor Belt Based on Machine Vision

摘要: 由于基于机器视觉输送带检测中输送带运动速度快、上下输送带间隔距离近(物距小)、采集范围大(视场大)，造成图像信息采集困难。本文提出了利用单个线阵CCD相机输送带图像的采集方法，该方法采集图像的视场角大，每个采集点只需要一个相机，网络传输方便，对网络传输要求低，图像处理简单，易于实现；利用工程光学理论，设计了单个相机的输送带图像在线采集系统，并对照明系统进行了设计。在实验室搭建了基于机器视觉的单个线阵CCD相机的输送带图像采集系统，进行了实验研究，结果表明该系统可实现高质量的输送带图像的采集，为研制基于机器视觉的输送带在线监测系统提供了基础。

Abstract: Mine conveyor belt has problems of fast speed, short distance and large field of view, and it is difficult to collect image information. Focusing on these problems, this paper proposes a method of using a single linear array CCD camera to collect the image of conveyor belt. When using this method, the field angle of the image capturing is large and each collection point only needs one camera. So that the transmission system is to facilitate the network, the network transmission requirements are low, and the image processing is simple and easy to implement. Under the theory of optical engineering, scheme of collecting the whole image of the conveyor belt by using a single linear array CCD camera was provided and the lighting system was specially designed. Belt image acquisition system on a conveyor belt was built in the laboratory and a large number of experiments were carried out. The experiments show that this system can collect high quality conveyor belt image information, and also can provide foundation for the development of the on-line image acquisition system of conveyor belt based on machine vision.