Traffic signal control system and traffic signal control method

Abstract

This invention discloses a traffic signal control system and a traffic signal control method. The traffic signal control system is used to control traffic signals and includes multiple image capturing devices and a computing device. The multiple image capturing devices are configured to acquire multiple raw images. The computing device is communicatively connected to the multiple image capturing devices and stores multiple reference images, wherein each reference image contains different identity labels, and each identity label corresponds to a different weight. The computing device is configured to perform the following operations: identify multiple different pedestrian profile images from the multiple raw images; compare each pedestrian profile image with the multiple reference images to filter out reference images that match each pedestrian profile image; classify the multiple pedestrian profile images into multiple identity groups according to the identity labels corresponding to the reference images that match each pedestrian profile image, wherein each identity group corresponds to a different identity label; and set the number of seconds for the traffic signal according to the weight corresponding to each identity group and the number of pedestrians in each identity group.

Claims

1. A traffic signal control system for controlling a traffic signal, the traffic signal control system comprising: Multiple image capturing devices are configured to generate multiple raw images; a computing device is communicatively connected to the multiple image capturing devices and stores multiple reference images, wherein each of the multiple reference images contains different identity labels, and each identity label corresponds to a different weight; the computing device is configured to perform the following operations: identify multiple different pedestrian profile images from the multiple raw images; compare each pedestrian profile image with the multiple reference images to filter out the reference images that match each pedestrian profile image; classify the multiple pedestrian profile images into multiple identity groups according to the identity labels corresponding to the reference images that match each pedestrian profile image, wherein each of the multiple identity groups corresponds to a multiple identity label; and set the duration of a traffic signal based on the weight corresponding to each identity group and the number of pedestrians in each identity group. The computing device is configured to overlay each of the pedestrian side profile images onto a plurality of the reference images, and to calculate a plurality of overlap ratios of each of the pedestrian side profile images onto the plurality of the reference images, wherein the reference image corresponding to the maximum value among the plurality of overlap ratios matches the pedestrian side profile image.

2. The traffic signal control system as claimed in claim 1, wherein one of the plurality of reference graphics includes a first feature graphic and a second feature graphic, the second feature graphic being connected to a front side of the first feature graphic, a first distance being between a top of the second feature graphic and a top of the first feature graphic, and a second distance being between a bottom of the second feature graphic and a bottom of the first feature graphic.

3. The traffic signal control system as claimed in claim 1, wherein one of the plurality of reference graphics includes a first feature graphic and a second feature graphic, the second feature graphic being connected to a rear side of the first feature graphic, a third distance being between a top of the second feature graphic and a top of the first feature graphic, and a fourth distance being between a bottom of the second feature graphic and a bottom of the first feature graphic.

4. The traffic signal control system as claimed in claim 1, wherein one of the plurality of reference graphics includes a first feature graphic and a second feature graphic, the second feature graphic being connected to a rear side of the first feature graphic, and a bottom of the second feature graphic being on the same horizontal plane as a bottom of the first feature graphic.

5. The traffic signal control system as claimed in claim 1, wherein one of the plurality of reference graphics includes a first feature graphic and a second feature graphic, the second feature graphic being connected to a front side of the first feature graphic, and a bottom of the second feature graphic being on the same horizontal plane as a bottom of the first feature graphic.

6. The traffic signal control system as claimed in claim 1, wherein one of the plurality of reference graphics is a square.

7. The traffic signal control system of claim 1, wherein the computing device includes a plurality of edge servo devices and a cloud servo device, the plurality of edge servo devices being electrically connected to a plurality of image capturing devices, the cloud servo device being network-connected to the plurality of edge servo devices, each edge servo device being configured to compare each pedestrian profile image with a plurality of reference images to find a reference image that matches each pedestrian profile image, each edge servo device being configured to classify the plurality of profile images into a plurality of identity groups based on the identity label corresponding to the reference image that matches each pedestrian profile image, the cloud servo device being configured to calculate the number of seconds for each group based on the weight corresponding to each identity group and the number of pedestrians, and the cloud servo device being configured to designate the maximum value among the plurality of group seconds as an adjustment number of seconds and transmit the adjustment number of seconds to the traffic signal.

8. The traffic signal control system of claim 7, wherein the edge servo device includes an edge processing circuit and an edge storage circuit, the edge storage circuit storing a probability judgment model and a plurality of the reference images, the edge processing circuit accessing the edge storage circuit to execute the probability judgment model, the probability judgment model being configured to compare each of the pedestrian profile images with the plurality of the reference images to find the reference image that matches each of the pedestrian profile images, the probability judgment model being configured to classify the plurality of feature images into a plurality of identity groups based on the identity label corresponding to the reference image that matches each of the pedestrian profile images.

9. The traffic signal control system of claim 8, wherein the cloud server includes a cloud processing circuit and a cloud storage circuit, the cloud storage circuit storing a second calculation module, the cloud processing circuit accessing the cloud storage circuit to execute the second calculation module, the second calculation module being configured to calculate the group second based on the weight corresponding to each of the identity groups and the number of pedestrians, the second calculation module being configured to designate the maximum value among the plurality of group second counts as an adjustment second count and transmit the adjustment second count to the traffic signal.

10. The traffic signal control system of claim 1, wherein the computing device is configured to calculate a group second based on the weight corresponding to each of the identity groups and the number of images in each of the identity groups, the computing device is configured to designate the maximum value among a plurality of the group second counts as an adjustment second count, and the computing device is configured to transmit the adjustment second count to the traffic signal to set the second count of the traffic signal equal to the adjustment second count.

11. A traffic signal control method, comprising: A computing device is configured to create multiple reference images, each of which contains multiple different identity labels and each identity label corresponds to multiple different weights; multiple image capturing devices are configured to generate multiple raw images; the computing device identifies multiple different pedestrian side profile images from the multiple raw images; the computing device compares each pedestrian side profile image with the multiple reference images to filter out the reference images that match each pedestrian side profile image; the computing device classifies the multiple pedestrian side profile images into multiple identity groups based on the identity labels corresponding to the reference images that match each pedestrian side profile image, each of the multiple identity groups corresponds to multiple identity labels; the computing device sets the duration of the traffic signal based on the weights corresponding to each identity group and the number of pedestrians in each identity group. The computing device filters out the reference graphics that match each of the pedestrian side profile graphics by: the computing device overlaying each of the pedestrian side profile graphics onto a plurality of the reference graphics; the computing device calculating a plurality of overlap ratios of each of the pedestrian side profile graphics onto the plurality of the reference graphics, wherein the reference graphic corresponding to the maximum value among the plurality of overlap ratios matches the pedestrian side profile graphics.

12. The traffic signal control method of claim 11, wherein the computing device includes a plurality of edge servers and a cloud server, each edge server comparing each pedestrian profile image with a plurality of reference images to find a reference image that matches each pedestrian profile image, each edge server being configured to classify the plurality of pedestrian profile images into a plurality of identity groups based on the identity label corresponding to the reference image that matches each pedestrian profile image, the cloud server calculating a group second based on the weight corresponding to each identity group and the number of pedestrians, the cloud server specifying the maximum value among the plurality of group second counts as an adjustment second count and transmitting the adjustment second count to the traffic signal.

13. The traffic signal control method of claim 11, wherein the computing device calculates a group second based on the weight corresponding to each of the identity groups and the number of images in each of the identity groups, the computing device designates the maximum value among a plurality of group seconds as an adjustment second, and the computing device is configured to transmit the adjustment second to the traffic signal so as to set the second of the traffic signal equal to the adjustment second.

14. The traffic signal control method of claim 11, wherein one of the plurality of reference graphics includes a first feature graphic and a second feature graphic, the second feature graphic being connected to a front side of the first feature graphic, a first distance being between a top of the second feature graphic and a top of the first feature graphic, and a second distance being between a bottom of the second feature graphic and a bottom of the first feature graphic.

15. The traffic signal control method of claim 11, wherein one of the plurality of reference graphics includes a first feature graphic and a second feature graphic, the second feature graphic being connected to a rear side of the first feature graphic, a third distance being between a top of the second feature graphic and a top of the first feature graphic, and a fourth distance being between a bottom of the second feature graphic and a bottom of the first feature graphic.

16. The traffic signal control method of claim 11, wherein one of the plurality of reference graphics includes a first feature graphic and a second feature graphic, the second feature graphic being connected to a rear side of the first feature graphic, and a bottom of the second feature graphic being on the same horizontal plane as a bottom of the first feature graphic.

17. The traffic signal control method of claim 11, wherein one of the plurality of reference graphics includes a first feature graphic and a second feature graphic, the second feature graphic being connected to a front side of the first feature graphic, and a bottom of the second feature graphic being on the same horizontal plane as a bottom of the first feature graphic.

18. The traffic signal control method as claimed in claim 11, wherein one of the plurality of reference graphics is a square.

Images