Intelligent traffic early warning method and system based on multi-modal visual tracking

You are here: Homepage > Search

Intelligent traffic early warning method and system based on multi-modal visual tracking

Free access

WANG FENGJU / LIANG QINJIAN / HUO JIANJIE

The invention discloses an intelligent traffic early warning method based on multi-modal visual tracking, and relates to the technical field of traffic early warning, and the method comprises the steps: obtaining the positioning information of a rescue destination, and determining a real-time idle coordinate set of a rescue vehicle; the rescue destination positioning information serves as a center, a rescue vehicle alternative coordinate set is screened and determined, a multi-modal sensor is used for data collection, and a multi-modal collection database is obtained; preprocessing the multi-modal acquisition database to obtain a multi-modal acquisition calibration database; performing feature extraction on the multi-modal acquisition calibration database; according to the method, the multi-modal characteristic indexes are subjected to characteristic fusion splicing, a traffic scene sensing model is constructed in combination with a rescue vehicle alternative coordinate set to determine a rescue vehicle target coordinate, and an emergency rescue task is issued to a vehicle-mounted person, so that the problem of relatively long time required by a rescue distance due to the fact that rescue work cannot be pre-warned in advance in the prior art is solved; the technical effect of improving the rescue efficiency is achieved.

本发明公开了基于多模态视觉跟踪的智能交通预警方法，涉及交通预警技术领域，方法包括：获取救援目的地定位信息，确定救援车实时空闲坐标集；将救援目的地定位信息作为中心，筛选确定救援车备选坐标集并使用多模态传感器进行数据采集，获取多模态采集数据库；对多模态采集数据库进行预处理，获取多模态采集校准数据库；对多模态采集校准数据库进行特征提取；将多模态特征指标进行特征融合拼接，结合救援车备选坐标集构建交通场景感知模型确定救援车目标坐标，并向随车人员发布紧急救援任务，解决了现有技术中存在的救援工作由于无法提前进行预警而导致救援路程所需时间较长问题，实现了提高救援效率的技术效果。