A detection method and system for an intelligent traffic electronic prompt device based on multi-feature vision

Abstract

The invention discloses a detection method and system for an intelligent traffic electronic prompting device based on multi-feature vision, which includes a video stream collection unit to collect road videos; a preprocessing unit converts the collected videos into models that are easy to analyze, transmit and further process The input unit transmits the traffic road data to the model; in the detection unit, the target detection model extracts the features of the data in the input model, outputs the feature map, and the semantic segmentation model classifies the feature points of the feature map and predicts the result; The output unit outputs the results; the voice reminder unit voice broadcasts the results. The present invention uses the idea of ​​combining target detection and semantic segmentation to accurately target the accuracy of target detection to specific feature points, which improves the detection accuracy, uses excellent network models to automatically extract high-level semantic information from original images, and improves feature extraction efficiency. , reduce labor costs, so that the driver has a good sense of experience and comfort.

Description

technical field [0001] The invention belongs to the technical field, and in particular relates to a detection method and system for an intelligent traffic electronic prompt device based on multi-feature vision. Background technique [0002] In the current real-time map navigation scenario, real-time reminders for different types of traffic electronic reminder devices are crucial in regulating the driving behavior of vehicle owners and ensuring user travel. However, various electronic prompting devices are distributed in the streets and alleys of the road. Such objects generally have small targets and are not easy to find. It is difficult to identify through conventional data collection and human means, which increases additional costs for the realization of navigation planning. In the prior art, the method of prompting about the electronic device is generally through the driver's visual inspection, but this method not only increases the difficulty of the driver's operation, ...

AI Technical Summary

Problems solved by technology

However, the efficiency and accuracy of existing target detection algorithms are low, which easily leads drivers to make wrong judgments on road conditions.

Images