Hoping I can build or buy.
Ok I'm getting a few years on my old eyes and the red/ yellow sometimes blended together on certain types of traffic lights. I am wanting to put and in car sensor that can sense the traffic light colors and output/trigger an led light of the same color. Needs to be on a timer so it goes off after a few seconds. Just enough to get my attention in the car.
Any ideas ?? Or someone that does this type of work.
Thanks

 

Yes, it can be done but it is not a simple task. Give this a read. The link is just an overview but autonomous cars need to read traffic lights. I am 73 and come the day I can no longer read traffic lights I surrender my truck keys before I kill myself or more important someone else's loved one.

Ron

 

If you cannot distinguish the red and yellow of a traffic light, you should not be driving.

 

I think a camera is the only answer. I don't think you can buy such an item, and to say it is not a simple task to create one is a huge understatement.

Tesla and Google, amongst others, do this kind of work! Perhaps you need to buy one of their cars.

 

Well shoot hoping it wasn't something to difficult. And for the ones wondering. Reds and yellows aren't really needed. Green is go.

 

Stop is way more important.

 

Hoping I can build or buy.
Ok I'm getting a few years on my old eyes and the red/ yellow sometimes blended together on certain types of traffic lights. I am wanting to put and in car sensor that can sense the traffic light colors and output/trigger an led light of the same color. Needs to be on a timer so it goes off after a few seconds. Just enough to get my attention in the car.
Any ideas ?? Or someone that does this type of work.
Thanks

Have your eyes checked. It is very likely that you have cataracts.

 

I imagine a camera would sense brake lights and some turn signals, enough to be impractical.

 

Funny, the green looks almost as white as a street light. Fortunately it's not in the Red/Green zone of my vision. I have a hard time distinguishing one from the other. Except for the fact that the "Green" of a traffic light is so much whiter than the red. Go figure.

If I couldn't tell color, at least I know that red is on the top and yellow is in the middle. There are plenty of totally color blind people driving today. Where it can get tricky is those traffic lights that are mounted horizontally. I know the yellow is in the middle. Just where the red is - that I don't know for sure. At least I can still distinguish the difference between a red traffic light and a green one.

 

If you cannot distinguish the red and yellow of a traffic light, you should not be driving.

color blond people are allowed to drive, they just know which position the lights are, red is top and yellow is middle.

 

Hoping I can build or buy.
Ok I'm getting a few years on my old eyes and the red/ yellow sometimes blended together on certain types of traffic lights. I am wanting to put and in car sensor that can sense the traffic light colors and output/trigger an led light of the same color. Needs to be on a timer so it goes off after a few seconds. Just enough to get my attention in the car.
Any ideas ?? Or someone that does this type of work.
Thanks

well if you want to buy, buy a tesla, they sense traffic lights.

 

This open source project using TensorFlow and Python looks very promising. It would be relatively trivial for someone with experience in systems administration and programming languages since it is so well documented but I am afraid it would not be too easy for a neophyte with no experience in any aspect.

 

color blond people are allowed to drive, they just know which position the lights are, red is top and yellow is middle.

Well, the color blind person may well have normal acuity. If his vision has deteriorated to to the point he can’t distinguish the colors, I doubt that he has the acuity required. I have never heard of losing color vision with age. I think @MrChips nailed it with cataracts, they cause loss of contrast and acuity.

 

I imagine a camera would sense brake lights and some turn signals, enough to be impractical.

Every modern vehicle that has automatic high beams can distinguish between red traffic lights and tail lights just fine.

So all the hardware is there, all the software is there, but no way no how any car manufacturer would take on the liability of telling any driver if the light is green, yellow or red.

Toyota was in a world of hurt because drivers could not keep their floor mats in place and blamed it on "unintended acceleration". Can you imagine the number of lawsuits when people start claiming their car said the light was green and it was actually red?

 

I think a camera is the only answer. I don't think you can buy such an item, and to say it is not a simple task to create one is a huge understatement.

Tesla and Google, amongst others, do this kind of work! Perhaps you need to buy one of their cars.

Maybe a system could be developed that a traffic light changes color AND sends an RF signal that a car sensor would listen to from a 1/4 mile range and alert a driver to the light change, and if needed, implement auto braking prior to arriving at an intersection.

 

Maybe a system could be developed that a traffic light changes color AND sends an RF signal that a car sensor would listen to from a 1/4 mile range and alert a driver to the light change, and if needed, implement auto braking prior to arriving at an intersection.

Radio isn’t a good way to deal with this at all. It would be impossible for the vehicle to know which of the several possible signals is sending the data. If you used highly directional signals it might be workable but that means complicated antennas and adding radios to traffic signals. It just won‘t scale.

But, there is already an EMR emitter on-board: the LEDs that show the signal. It would be quite possible to use something like Li-Fi, or Li-Fi itself, to have the active signal send information about its status. Even how long the signal will remain in the state it is in (though this could have undesirable side effects if people tried to game the lights). If using the existing LED array was problematic for some reason, a secondary array of specialized emitters (e.g.: narrowly focused IR) could be used.

A camera-based system would be able to determine a light transmitting a “signature“ is a traffic signal, and what state it is in. It could even send it‘s GPS coordinates so the SatNav system could be used to confirm the identity and even place it on the map in realtime. If needed (because of, say, insufficient refresh rate), a secondary sensor could be used to decode the data after the camera recognizes it is a candidate.

 

Certainly it will take a fairly high resolution camera, followed by software to recognize the shape of a traffic light from among all of the other lights along the road. Then discerning red should not be too difficult.

BUT it was a driver who did not see the traffic light at an adequate distance, and so left the road and bashed in the side of my friend's restaurant and narrowly missed killing six of my friends working there, who should not have been driving, because of not seeing the traffic light. So I seriously suggest that the TS either stop driving or get a lens replacement for their eyes. The vision improvement is quite impressive. Avoid killing people by accident.

 

Radio isn’t a good way to deal with this at all. It would be impossible for the vehicle to know which of the several possible signals is sending the data. If you used highly directional signals it might be workable but that means complicated antennas and adding radios to traffic signals. It just won‘t scale.

But, there is already an EMR emitter on-board: the LEDs that show the signal. It would be quite possible to use something like Li-Fi, or Li-Fi itself, to have the active signal send information about its status. Even how long the signal will remain in the state it is in (though this could have undesirable side effects if people tried to game the lights). If using the existing LED array was problematic for some reason, a secondary array of specialized emitters (e.g.: narrowly focused IR) could be used.

A camera-based system would be able to determine a light transmitting a “signature“ is a traffic signal, and what state it is in. It could even send it‘s GPS coordinates so the SatNav system could be used to confirm the identity and even place it on the map in realtime. If needed (because of, say, insufficient refresh rate), a secondary sensor could be used to decode the data after the camera recognizes it is a candidate.

After I posted about RF signals, I realized that directional sending would be a hassle.

If cameras were used, and one is following a large truck, most of the visual signal available to a camera would be blocked. Much like current conventional driving practices and experiences. That might really mean one is following too closely...not sure

 

No DIY solution will help you pass your next vision test. Get your cataracts done. Everyone I know that had it done laments waiting so long. Ditto for hip replacements.

 

One more consideration relative to the communication with vehicles as touted by some fools: If it is used for safety then it becomes a life-critical function, and the reliability requirements are rather demanding. And that would include every car and truck. So suddenly the cost to everybody becomes much more. Not just a bit more, but at least doubled. Do most folks want to pay dearly for a system that they do not need???