Vehicle detection method

Abstract

A vehicle detection method comprises two parts: training of cascade multiple classifier and using the classifiers to carry out vehicle detection. The method is characterized by: using a LBP operator representation pixel level characteristic and a LBP histogram representation pixel block characteristic, and using a LBP spatial histogram with spatial information to characterize spatial structure information of the vehicle; using the LBP histogram characteristic to train a linear SVM classifier, using a boost method to train a second layer HIKSVM classifier, and detecting the vehicle according to a final classifier result. The LBP spatial histogram characteristic is used to characterize the vehicle information so that a vehicle outline can be rapidly characterized. Certain tolerance is possessed to a variety of an intra-class characteristic and strong distinguish is possessed to the variety of the outer class. The linear SVM classifier is used so that purposes of rapid training and rapid detection can be achieved. And real-time performance can be achieved too. The boost training method is used to train the second layer HIKSVM classifier so that a detection rate of the vehicle can be increased.

Description

Technical field [0001] The present invention relates to the technical field of video surveillance, in particular to a vehicle detection method based on LBP spatial histogram features and cascaded multi-classifiers, suitable for different scenes and different lighting, and achieving real-time performance. Background technique [0002] In the existing vehicle detection methods, one implementation method is vehicle detection based on background modeling. This method uses a certain strategic background modeling method to extract the foreground motion area, and then further determines whether there is a vehicle, but the method is illuminated The impact of the vehicle is relatively large, which belongs to the earlier primary detection method; the other implementation method is based on the two-stage method of vehicle hypothesis generation and hypothesis verification, which is based on certain characteristics (such as the shadow under the vehicle, the vehicle’s Symmetry, gradient, edge,...

AI Technical Summary

Problems solved by technology

[0002] In the existing vehicle detection methods, one implementation method is vehicle detection based on background modeling. This method uses a background modeling method with a certain strategy to extract the foreground moving area, and then further judges whether there is a vehicle. However, this method is affected by light. has a greater impact, and belongs to the earlier primary detection method; another implementation method is based on the two-stage method of vehicle hypothesis generation and hypothesis verification, which is based on certain features (such as the shadow under the vehicle, the vehicle's Symmetry, gradient, edge, etc.) to generate vehicle assumptions, and then extract features to verify the assumptions. This method has high requirements for vehicle assumption generation, and vehicles that do not meet the assumption generation conditions (such as vehicle congestion, cloudy, etc.) will be missed. check

In addition, the method based on histogram of gradients (HOG) has been successfully applied to pedestrian detection, but its running speed is difficult to achieve real-time performance. Corresponding hardware cost

Images