YOLOv5在针织品缺陷检测中的应用

The Application of YOLOv5 in Defect Detection of Knitted Fabrics

针织品的质检通常由人工完成,对瑕疵的判断易受各种主观因素的影响。随着计算机技术的高速发展,CNN、RNN等网络为其智能检测提供了新的启示。YOLOv5作为一种实时检测方法,可将物体检测转化为一种回归问题,从而简化检测过程。因此,针对基于计算机视觉技术质检,本文通过对各种改进算法进行分析,并将质检领域的研究现状进行对比,从而为该技术的研究提供参考。

The quality inspection of knitted goods is usually done manually,and the judgment of defects is easily affected by various subjective factors.With the rapid development of computer technology,networks such as CNN and RNN provide new enlightenment for its intelligent detection.YOLOv5,as a real-time detection method,can transform object detection into a regression problem,thus simplifying the detection process.Therefore,aiming at the quality inspection based on computer vision technology,this paper analyzes various improvement algorithms and compares the research status in the field of quality inspection to provide reference for the research of this technology.