Traffic light signal cycle extracting method based on data of automobile data recorder

Abstract

The invention discloses a traffic light signal cycle extracting method based on the data of an automobile data recorder, which includes the following steps: reading video data; inputting an image; transforming the image from an RGB color space to an HSV color space; carrying out segmentation according to a component v and excluding the interference of other factors to get the state of signal lights; drawing the state of the signal lights to a specific coordinate system to get time-varying state diagrams of the signal lights; grounding the time-varying state diagrams through a Mean Shift clustering method, and excluding the interference of discrete points; and calculating the signal cycle according to the data characteristics of each group. The advantages are as follows: (1) there is no need to record signal lights by meter on site, and the risk of data collection is reduced; (2) there is no need to copy time, state and other data on paper, and the workload of collection is reduced; (3)the time cost and labor cost of data collection are low; and (4) different cycles of the same signal light can be obtained.

Description

technical field [0001] The invention relates to the technical field of intelligent transportation, in particular to a method for extracting traffic light signal periods based on driving recorder data. Background technique [0002] Signal-controlled intersections are an important part of the urban road network. Generally, various signal lights are used to control the passing time, cycle and other information of each direction of traffic. The identification and analysis of the signal light cycle at signal-controlled intersections is of great significance to the fields of automatic driving and intelligent transportation. In foreign countries, as part of government information, signal cycle data has corresponding public access channels. In China, because the signal light system is set up by different units and lacks a unified information management approach, there is no complete and unified city signal light information. Therefore, the current signal light data collection meth...

AI Technical Summary

Problems solved by technology

In China, because the signal light system is set up by different units and lacks a unified information management approach, there is no complete and unified urban signal light information

Therefore, the current signal light data collection method mainly collects signal light information manually, which takes a long time period, consumes a lot of manpower and material resources, and the collection process has certain risks.

Images