传送带偏移检测技术研究及FPGA实现

Research on Conveyor Belt Offset Detection Technology and FPGA Implementation

摘要:

为了解决送料传送带在运行过程中发生的横向偏移和纵向偏移所导致的材料加工精度降低等问题，设计了一种基于改进的Hough直线检测算法的传送带偏移检测系统。本系统采用FPGA作为处理单元，对CMOS图像传感器拍摄的实际传送带图像进行滤波、边缘检测、二值化等图像的预处理，利用以太网传输方式把预处理后的图像信息传送给上位机，在上位机中进行Hough直线检测算法处理，获得传送带的偏移状态。通过实验，系统能够实时地检测出倾斜角精度最高0.245°的倾斜偏移和横向8个像素的偏移。

Abstract:

In order to solve the problems of material processing accuracy reduction caused by lateral and longitudinal migration during the operation of the feeding belt, a conveyor migration detection system based on the improved Hough straight line detection algorithm was designed. A FPGA chip is used as the processing unit to filter, edge detection, binarization and other image preprocessing of the actual conveyor belt image taken by the CMOS image sensor in this system. The pre-processed image information is transmitted to the upper computer using the Ethernet transmission method, and the Hough line detection algorithm is processed in the upper computer to obtain the offset state of the conveyor belt. Through experiments, the system can detect in real time tilt offsets with an inclination angle greater than 0.245 degrees and offsets of 8 pixels in the lateral direction.