Pedestrian red light face recognition system is set up at the intersection of an electronic display and face recognition camera, when the crosswalk traffic light from green to red, the system will immediately sound a reminder: "Now is a red light, please do not run the red light." If the pedestrian insists on running the red light, when he or she walks to the middle of the crosswalk, the equipment will simulate the sound of brakes, so as to warn the pedestrian. Not only that, the system will also carry out face-to-face recognition of pedestrians who run the red light, and capture, take evidence of three photos before, during and after running the light.

The first: crosswalk passes when the display is red, pedestrians just ready to run the red light, the system will take the first photo;

The second: when pedestrians ignore the sound reminder, continue to run the red light, the photographic equipment will continue to take the second photo;

The third: when pedestrians walk to the bottom of the camera, the exposure equipment will emit a sound similar to the sound of braking sharply as well as the vehicle collision to deter pedestrians, hopefully warning the person to return to the intersection, while again taking a photo of the pedestrian running the red light.

At this point, the system has recorded the pedestrian running the red light. The system will intercept a close-up picture of the pedestrian from the photo taken, to be able to see and analyze the full picture of the pedestrian, and these photos of pedestrians running the red light will be scrolling on the big screen at the scene until the person who ran the red light takes the initiative to contact the traffic police, agreeing to accept the punishment and remove the photo. As for those who are indifferent, the traffic police will also take the initiative to contact the person concerned through the big data system. The system is also said to be connected to the household registration system to directly access the household registration information of red-light runners, and these red-light running records will also be included in the individual's credit system.

Problems with red-light running face recognition systems

1, voice warnings are dangerous

In the process of pedestrians running a red light a **** there are two places where voice warnings will appear, one in the traffic light from green to red, the system will remind pedestrians not to run the red light; the second is in the middle of the pedestrians into the crosswalk, the system will simulate the brakes to stimulate the sound of pedestrians.

This kind of voice warning has a certain effect on stopping some pedestrians from running the red light, after all, many people will run the red light just because of the momentary slackness in their hearts, and once there is someone to remind them, many people will be able to pull back from the brink.

But there are some problems with this voice warning. First, there's the braking warning when a pedestrian reaches the middle of a crosswalk. Because, this time is the red light time, pedestrians standing in the crosswalk is a certain danger, and this kind of simulation of the sound of the brakes will scare the pedestrians, and even more increase their danger, this is one of them; Secondly, the voice of the prompt in fact can play a big role, in fact, it is difficult to say.

2, photo exposure invasion of privacy

Intelligent pedestrian red light forensics system is not only able to forensics, but also will be able to red light trespassers in the big screen exposure in real time, to play a punitive and deterrent effect. In the eyes of law enforcement, this system can force the public to consciously abide by traffic regulations, but in the eyes of some members of the public, this practice is suspected of violating personal privacy.

At present, face recognition has been widely used in finance, justice, military, public security, border control, government, aerospace, electric power, factories, education, medical care, and many enterprises and institutions, etc., and its rich application scenarios bring huge market potential to face recognition. With the gradual increase in application scenarios, the layout of the face recognition ecosystem is also richer, from consumer electronics to security, finance and so on.

In the face of the fire face recognition market, focusing on artificial intelligence technology scene innovation and landing application of gravity interconnection, and byte jumping's enterprise-oriented digital services and intelligent technology brand volcano engine signed a cooperation agreement in Beijing Haidian District. The two sides based on artificial intelligence algorithm mall - titanium spirit AI algorithm market, around face recognition, gesture recognition, human body key point detection and other computer vision (CV) cutting-edge technology direction, combined with practical application scenes and data to carry out in-depth cooperation.

Along with the "new infrastructure" in artificial intelligence, Internet of Things, 5G and other aspects of the continued investment, it can be predicted that in the next ten years, we will accelerate into the "visual Internet of Things era", face recognition as the main technology direction of the field of computer vision, in the "era of visual Internet of Things", face recognition as the main technology direction of the field of computer vision, face recognition as the main technology direction of the field of computer vision. Face recognition, as the main technology direction in the field of computer vision, will definitely get explosive growth in the "visual Internet of Things era".