Expressway agglomerate fog early warning system based on deep fusion network

Abstract

The invention discloses an expressway agglomerate fog early warning system based on a deep fusion network. The method comprises the steps of 1, obtaining a to-be-detected image; 2, performing high-frequency filtering to obtain a frequency domain characteristic graph; 3, extracting a saturation component of the to-be-detected image as a saturation graph; 4, performing spectral transformation on theto-be-detected image to obtain a spectrogram; 5, respectively sending the frequency domain characteristic graph, the spectrogram and the original graph into a convolutional neural network; 6, splicing and fusing the three extracted characteristics; 7, carrying out classification to obtain classification grade information of fog; 8, constructing a training sample; 9, training a deep learning network based on a CNN; 10, for the to-be-detected image, judging the visibility level of the current camera monitoring road section according to the classification result; and 11, sending a low-visibilityalarm and an agglomerate fog early warning to the traffic management department of each road section. According to the invention, the advantage of massive monitoring of an expressway is more effectively utilized, and timely early warning is carried out.

Description

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine

Login to View More

AI Technical Summary

Benefits of technology

The technology described in this patented allows for accurate detection of raindrops from images captured at different locations over time without sacrificing detail or identifying specific areas within them. It also improves upon existing methods such as histogram analysis and machine vision techniques while still maintain their effectiveness across varying weather conditions. Overall, these technical improvements improve the performance of models trained through neural networks (ANN).

Problems solved by technology

This patented describes how we want to monitor and warn clouds from haze when there's no clear view through windows like roads during winter months due to misty precipitation. Current methods involve manually checking out vehicles overlapping areas where they may pose safety concerns while driving. Additionally, current systems use multiple sensors (photographic elements) but only collect scenes along one direction - either side upwards or towards them.

Images