Black smoke vehicle recognition method based on Gaussian mixture and autoregressive moving average model

Abstract

a black smoke vehicle recognition method based on Gaussian mixture and an autoregressive moving average model comprises the following steps: (1) detecting moving objects from road surveillance video using a Gaussian mixture model; (2) extracting three kinds of features of a vehicle key area, including Haar-like features, co-occurrence matrix gradient direction histogram features and local binary mode Fourier histogram features; (3) using an ARMA model to model the continuous multi-frames of each feature, and obtaining three different models; (4) for the new vehicle target, using the three models to classify the three features extracted from the key areas of the vehicle. Combined with the classification results of different features and the comprehensive analysis of continuous multi-frames,the method can judge whether there are smoky vehicles in the current video segment. The invention can greatly save manpower and financial resources consumed in a traditional method, is conducive to obtaining and preserving evidence, does not affect normal traffic, and can effectively improve law enforcement efficiency.

Description

technical field [0001] The invention relates to the technical fields of computer vision and pattern recognition, in particular to a smoky car recognition method based on a Gaussian mixture and an autoregressive sliding average model. Background technique [0002] Diesel vehicles with serious pollution are also called smoky vehicles. Usually there will be thick black smoke at the exhaust holes. The pollution has always been the focus of motor vehicle pollution control. The smoky vehicles driving on the road are found in time and the environmental protection Further processing by the department will help reduce motor vehicle pollution, improve air quality, and reduce its harm to the human body. [0003] At present, many cities still use road inspections, public reports and manual video surveillance to detect smoky vehicles on the road, which is time-consuming and laborious, and will affect normal traffic operations, which is not conducive to the preservation of illegal evidenc...

AI Technical Summary

Problems solved by technology

[0003] At present, many cities still use road inspection, public reporting and manual video surveillance to detect smoky vehicles on the road, which is time-consuming and laborious, and will also affect normal traffic operations, which is not conducive to the preservation of illegal evidence of related vehicles

Images