Coat and cap characteristics and attitude detection-based traffic police detection method and system

Abstract

The invention relates to a coat and cap characteristics and attitude detection-based traffic police detection method and system. The system comprises an offline training construct classifier, an online real-time detection traffic police and an attitude thereof, wherein an offline part extracts a pedestrian hog characteristic, a cap hog characteristic, a fluorescent clothes hog characteristic, a construct pedestrian classifier, a cap classifier and a fluorescent clothes classifier through a lot of collected sample images respectively; and an online part collects an original image in real time and judges whether a pedestrian, namely the traffic police who wears the cap and fluorescent clothes at the same exists in the original image or not by the pedestrian classifier, the cap classifier, the fluorescent clothes classifier and various combinations, then judges the attitude of the traffic police, and prompts a driver to help the driver correctly understand the attitude of the traffic police. The detection method is broadly divided into two processes of accurate matching and fuzzy matching; the two processes can be corroborated by each other; the traffic police and the attitude thereof can be accurately and quickly detected; and the coat and cap characteristics and attitude detection-based traffic police detection method and system have important guiding significance in active driving.

Description

technical field [0001] The invention belongs to the field of active driving, in particular to a traffic police detection method and system based on clothing features and posture detection. Background technique [0002] In recent years, the development of autonomous driving technology in automobiles has been in full swing, and various automakers and technology companies have launched their own autonomous driving test concept cars. These experimental concept cars use pedestrian detection, vehicle detection, traffic signal detection, GPS positioning and map search, path planning and other technologies to guide them to drive autonomously, or even unmanned. These experimental cars can control and drive the car almost completely without human operation under some relatively common road conditions. But in some special traffic scenarios, it is the traffic police who direct the behavior of vehicles. However, the current self-driving vehicles have not actively responded to the traff...

AI Technical Summary

Problems solved by technology

However, the current self-driving vehicles have not actively responded to the traffic police.

Images