Axis type recognition method based on deep convolutional neural network

Abstract

The invention relates to an axis type identification method based on a deep convolutional neural network. According to the method, a target axle image is segmented from a vehicle, a deep convolutionalneural network is established to extract features from the axle image, feature fusion and expansion are carried out, regression training is carried out on the expanded features through a Softmax classification layer, and classification and identification of axle features are realized. According to the axis type recognition method based on the deep convolutional neural network, non-contact recognition is achieved by conducting image processing and recognition on the video images collected by the camera installed in the specific area, damage to the road surface is avoided, and the device is simple and easy to maintain.

Description

technical field [0001] The invention belongs to the technical field of image recognition, in particular to an axis type recognition method based on a deep convolutional neural network. Background technique [0002] Axle type recognition is the most important part of highway overrun detection. The traditional axle type classification is to identify the wheel axle through the sensor laid on the road or the grating erected, but this method is difficult and expensive to maintain the equipment. [0003] For example, the shaft type identification is realized through a dynamic weighing detection system. This weighing system converts the pressure it bears into a voltage signal through a sensor laid on the road surface, and judges the number of shafts and shaft type according to the output waveform. However, the realization of this method requires the laying of sensors on the road surface, and the equipment maintenance is not easy, which will cause damage to the road surface; [00...

AI Technical Summary

Problems solved by technology

[0006] The invention relates to an axle type recognition method based on a deep convolutional neural network, which solves the problem of difficult maintenance and high cost in the traditional technology of identifying wheel axles through sensors laid on the road or gratings erected.

Images