SVM vehicle axle number detection method based on HOG features and MB-LBP features

Abstract

The invention discloses an SVM vehicle axle number detection method based on HOG features and MBLBP features. The method comprises the steps: (a) building a vehicle axle number and axle distance measurement system based on vision; (b) collecting sample images, wherein a tire image and a non-tire image are respectively taken as a positive sample and a negative sample; (c) extracting an HOG featurevector of the sample; (d) extracting an MB-LBP feature vector of the sample; (e) fusing the HOG feature vector and the MB-LBP feature vector of the sample into a final feature vector of the sample; (f) training the SVM by taking the final feature vector as a training sample of the SVM to obtain an SVM model for tire recognition; and (g) shooting a tire image of the vehicle to be detected, identifying the tire by using the trained SVM model, and counting the axle number of the vehicle. According to the method, the HOG features and the MB-LBP features are fused, the accuracy of feature description is enhanced, the real-time performance of a detection algorithm is improved, and the method can be used for metering the vehicle wheelbase.

Description

technical field [0001] The invention belongs to the technical field of computer vision, and in particular relates to an SVM vehicle axle number detection method based on HOG features and MB-LBP features. Background technique [0002] The number and type of axles is an important indicator to be considered in the weighing process of freight vehicles. In the process of weighing the vehicle, technicians should accurately identify the number and type of axles of the target vehicle, and use corresponding technical standards to measure whether the vehicle is overweight. At present, in most areas of our country, the number and type of axles of vehicles are still judged manually, which is inefficient and difficult to guarantee the accuracy of the results. Bring economic losses, but also bring serious threats to road transport traffic safety. Therefore, how to design a set of high-precision automatic axis detection system is an urgent problem to be solved. In the field of target re...

AI Technical Summary

Problems solved by technology

At present, in most areas of our country, the number and type of axles of vehicles are still judged manually, which is inefficient and difficult to guarantee the accuracy of the results. Bring economic losses, but also pose a serious threat to road transport traffic safety

Images