Non-motor vehicle driving identification method and device, computer and readable storage medium

Abstract

The invention discloses a non-motor vehicle driving identification method and device, a computer and a readable storage medium, and relates to the field of traffic transportation, and the method comprises the steps: making a captured picture into a data set, and carrying out the marking and training of a driver and a non-motor vehicle in the data set, and obtaining a driver detection model and a non-motor vehicle detection model; acquiring a driver object and a non-motor vehicle object; labeling the driver object to obtain a driver single-label data set and a driver multi-label data set; labeling non-motor vehicle objects to obtain a non-motor vehicle single-label data set and a non-motor vehicle multi-label data set; corresponding single-label and multi-label recognition models are trained for the driver single-label and multi-label data sets; performing corresponding single-label and multi-label identification models on the non-motor vehicle single-label data set and the non-motor vehicle multi-label data set; and combining the corresponding single-label identification model and the corresponding multi-label identification model to obtain a corresponding driver identification model and a non-motor vehicle identification model. The method has the advantages of accurate attribute recognition and high speed.

Description

technical field [0001] The invention relates to the field of transportation, in particular to a non-motor vehicle driving identification method, device, computer and readable storage medium. Background technique [0002] With the continuous development of artificial intelligence, its application in intelligent transportation has become increasingly mature, and the traditional control of motor vehicles can no longer meet the needs of relevant departments, especially with the rise of the food delivery industry, electric vehicles and bicycles, etc. Behaviors such as non-motor vehicles running red lights, driving motor vehicle lanes, and overloading are very common, so it is extremely important to detect the driving behavior of non-motor vehicles and automatically distinguish the attributes of the driver and the vehicle they are driving. [0003] At present, there are relatively few researches on the attribute recognition of people riding bicycles at home and abroad, and most of...

AI Technical Summary

Problems solved by technology

[0004] Aiming at the problem that the existing attribute recognition model cannot give consideration to both reasoning speed and recognition accuracy, the present invention provides a non-motor vehicle driving recognition method, device, computer and readable storage medium

Images