A smoky vehicle detection method based on pixel adaptive segmentation and Bayesian model

Abstract

The invention discloses a black smoke vehicle detection method based on pixel adaptive segmentation and Bayesian model. The black smoke vehicle detection method comprises the following steps: detecting a moving target by using a PBAS model, dividing an image into grids, and marking all small squares where a foreground target is located; for each foreground target grid, extracting POEM histogram orLDP histogram to characterize the spatial information, for each foreground target grid, extracting HOOF histogram or MOH histogram to characterize the temporal information, for each foreground targetgrid, extracting the STH feature and depicting the structure information, fusing different types of histogram features, using the characteristics of non-smoky vehicles being much higher than smoky vehicles in the actual scene to add a priori knowledge, classifying the small squares of each foreground target in each frame by Bayesian model, and comprehensively analyzing multi-frame recognition ofsmoky vehicles. The invention can automatically identify the black smoke vehicle from the traffic flow, improves the detection rate, reduces the false alarm rate, and has robustness to shadows.

Description

technical field [0001] The invention belongs to the technical field of pattern recognition and image processing, and relates to a method for detecting a smoky vehicle, in particular to a method for detecting a smoky vehicle based on pixel adaptive segmentation and a Bayesian model. Background technique [0002] Smoky vehicles are highly polluting vehicles. The country attaches great importance to the detection of smoky vehicles, which is of great significance to reducing motor vehicle pollution and improving air quality. [0003] At present, the research on smoky vehicle detection algorithm based on video analysis is still in its infancy, and there are not many inventions that can be referred to. end of 2016 et al. proposed a multi-camera-based vehicle exhaust analysis and black-smoky vehicle detection system for the first time. This method uses far-infrared cameras to locate the vehicle exhaust outlet, and determines the degree of exhaust pollution by analyzing the area a...

AI Technical Summary

Problems solved by technology

The method compensates The method uses multiple cameras, but there are still some problems. For example, the accuracy of Vibe background difference algorithm to detect vehicles needs to be improved. It only considers the extraction of a single Tamura feature, and does not consider the dynamic characteristics of black smoke, which is prone to false alarms.

In general, many methods ignore dynamic features, and shadows can easily cause false detection

Images