Diesel vehicle black smoke image recognition method and system and storage medium

Abstract

The invention discloses a diesel vehicle black smoke image recognition method and system and a storage medium. The method comprises the steps of: obtaining a diesel vehicle black smoke video, inputting the video into a black smoke image recognition model trained in advance, carrying out the testing, obtaining a dynamic black smoke segmentation video, and recognizing a black smoke ball. A black smoke image recognition model training step comprises the following steps: sample construction: data is a video for monitoring and shooting starting of a diesel vehicle at a traffic light intersection, discharged black air pollutants are recorded, and the video is divided into a plurality of frames; and network construction: two frames are randomly selected from [(c-1)*k, c*k) frames to consider the context relation with the current frame in each k frames, namely the system structure is divided into a first two-frame association information extraction module, namely CEM, and a current frame segmentation module CFSM, the (c*k)th frame being the currently to-be-processed frame. According to the deep learning method based on semantic segmentation, whether black smoke exists or not can be detected, the size and shape of discharged black smoke can be intelligently described in real time, and evaluation of the pollution degree of vehicles is facilitated.

Description

technical field [0001] The invention relates to the technical field of intelligent segmentation of diesel vehicle exhaust, in particular to a method, system and storage medium for recognizing black smoke images of diesel vehicles. Background technique [0002] The black smoke exhaust gas emitted by diesel vehicles contains about 200 different compounds, which is one of the main causes of fine particulate matter and photochemical smog pollution. Considering the mileage and emission coefficient of diesel vehicles comprehensively, the nitrogen oxide and fine particle emissions of a diesel vehicle that cannot meet the National III emission standards or lower emission standards are equivalent to the emissions of more than 200 National IV emission standard cars Sum. In view of the rapid increase in the number of motor vehicles on urban roads, the increasingly prominent exhaust pollution, and the increasing pressure of control, there is an urgent need for a traditional method of e...

AI Technical Summary

Problems solved by technology

The artificial feature extraction methods based on empirical thresholds and experimental statistical thresholds used in traditional smoke detection algorithms, such as artificially designing and modeling smoke color features, frequency domain features, texture features, etc., are difficult to perform detection tasks in such complex scenes

[0004] At present, all computing tasks of machine learning algorithms can only be deployed on the CPU, and the real-time performance is poor. It is difficult to perform specific optimization according to specific application scenarios and support GPU parallel computing acceleration.

Images