基于雾天交通标志识别算法的研究

Research on Algorithm of Traffic Sign Recognition Based on Fog

针对雾霾天气导致交通标志难以被准确识别的问题,本文提出在雾霾天气下快速实现交通标志检测识别算法的研究。通过全局直方图均衡、局部直方图均衡、单尺度Retinex和多尺度Retinex四种去雾算法实现含雾图像的去雾处理。采用YOLOv1目标检测算法对交通标志目标进行精准定位和检测。通过Alex-Net分类识别模型实现对LISA交通标志数据集中14类交通标志的识别。结果表明,本文的方法能使快速准确地对雾天交通标志进行检测、分类和识别。

Aiming at the problem that the traffic signs are difficult to be accurately recognized in foggy weather,this paper proposes a research on the rapid realization of traffic sign detection and recognition algorithm in haze weather.Firstly,four dehazing algorithms including global histogram equalization,local histogram equalization,single-scale Retinex and multi-scale Retinex are used to realize the dehazing process of the foggy image.Secondly,the YOLOv1 target detection algorithm is used to accurately locate and detect traffic sign targets.Finally,the recognition of 14 types of traffic signs in the LISA traffic sign dataset is realized through the Alex-Net classification recognition model.The results show,the method in this paper can quickly and accurately detect,classify and recognize foggy traffic signs.