The invention provides a vehicle detection method, a traffic flow statistical method and device, and an application vehicle detection model, the vehicle detection model comprises a backbone network, a fusion network and a prediction network, the method comprises the following steps: using the backbone network to extract multi-scale image features of a vehicle image, the backbone network comprising N C3 modules, the hierarchical structure of each C3 module is a Swin Transform structure, and N is an integer greater than or equal to 2; performing feature fusion on the multi-scale image features by using the fusion network to obtain a plurality of fusion features; and processing each fusion feature by using the prediction network to obtain a plurality of vehicle detection results. According to the method provided by the embodiment of the invention, the vehicles in the vehicle image can be accurately classified and tracked, and the traffic flow can be accurately counted.
本公开提供一种车辆检测方法、车流量统计方法及装置,应用车辆检测模型,所述车辆检测模型包括主干网络、融合网络和预测网络,所述方法包括:利用所述主干网络提取车辆图像的多尺度图像特征,所述主干网络含有N个C3模块,每个所述C3模块的层级结构为Swin Transformer结构,N为大于或等于2的整数;利用所述融合网络对所述多尺度图像特征进行特征融合,获得多个融合特征;利用所述预测网络对每个所述融合特征进行处理,获得多个车辆检测结果。本公开示例性实施例提供的方法,不仅可以对车辆图像中的车辆进行精准分类与跟踪,还可以准确的进行车流量的统计。
Vehicle detection method, traffic flow statistical method and device
一种车辆检测方法、车流量统计方法及装置
2023-05-12
Patent
Electronic Resource
Chinese
Traffic flow statistical method based on urban moving vehicle detection
European Patent Office | 2023
|Vehicle detection method, vehicle detection device, vehicle and traffic control device
European Patent Office | 2021
|European Patent Office | 2020
|