What sensors to use for project?

[jackbob]

Hello everyone! I have a friend working on a project who recruited me to work on the electronics aspect of the design and I have worked on many projects but an a bit puzzled with this one. Unfortunately he is currently in the process of obtaining a patent for the idea so I cannot speak the exact details of what I am trying to do but I do have a great analogy/representation of the goal I am trying to accomplish and I would greatly appreciate any suggestions.

Basically here is the idea. Imagine a basketball hoop without the net. What I am trying to do is sense when the basketball actually passes through the hoop. To make it a little more difficult the basketball would be about the size of a baseball relative to the hoop. So I need to detect something the size of a baseball passing through a basketball hoop. My best idea yet is to integrate infrared line break sensors across from each other and detect when any of the sensors detect something crossing their path. Is there any other sensors or methods that any of you can suggest?

Just when that seems like an easy goal... this is the hard part. Now imagine I have x number of baseballs say a reasonable number (2-20 balls). I want to detect exactly which ball passed through the hoop. Now again please keep in mind all this is an analogy, I am not actually trying to detect baseballs but similar objects. That said these balls can have electronics embedded within them in order to identify them. This is the part I need the real help with. How can I distinguish these balls from each other and determine which one passed through the hoop? Just so I don't seem like a lazy guy looking for answers I have thought of an idea but I am not certain how reliable it will be or if it is the best option. My best idea yet is to implant an RFID tag within the ball and have several readers in the hoop to detect the tag when it passes through. The problem with this is obvious. Imagine trying to read an RFID tag of a baseball falling through a basketball hoop. I have limited experience with RFID and am not certain if it will work for this application.

I thought about using RFID tags with reasonable rage (say 5 feet or so) this way the ball will be in rage of the reader long enough to be read. But I dont want the range to be too great to where it detects stray balls near the reader. Also a read tag will only be counted as long as the ball is detected by the earlier explained IR line break sensors (this almost ensures the ball that is read is the one that passed through the hoop).

Any suggestions or help would be greatly appreciated. Perhaps there are some sensors or techniques I am unaware of. The ideas I talked about are the best I have to accomplish this task. Also I would like to keep it relatively simple and not have crazy mechanisms like image detection and recognition software (I know its possible and could work but come on...) If there is any way my idea could be improved or made possibly more reliable I welcome your response! Thank you guys.

[JacquesBBB]

Depends on the speed, and on the budget.

You can paint the bals in dif colors. Put a high speed camera and fast recognition software.

[jackbob]

Also I would like to keep it relatively simple and not have crazy mechanisms like image detection and recognition software (I know its possible and could work but come on...)

[jackbob]

Thank you for the suggestion. I honestly don't think a camera recognition software will work well in the final product as one thing I did not mention but didn't think it would be relavent is that the "basketball hoop" position may not be constant which would introduce great difficulty in image tracking. It's position won't vary greatly but it won't be stable. As for the active RFID that seems like a great idea they appear to broadcast rather far though but it looks as if it could work. Like you said read the closest tag. I know this may seem rather silly but is it feasible to "overclock" a passive RFID tag reader? What I mean by that is amplify the signal it normally puts out to a passive RFID tag so that its distance may be a little greater. Because passive tags work on power transmitted by the reader would it be at app possible to have a few readers nearby and perhaps somehow amplified a bit? Then the problem I see with that is if a tag is read too close it could induce too high a voltage kn the passive tag and ruin it.

[han]

https://youtu.be/KAm7qAKAXwI

[jackbob]

Han, very interesting video thank you for the share I didn't know and never would of guees how those worked! They are very interesting but I think about it and I can't see exactly how this would identify which object passes through the hoop. It is an interesting way though to tell when an object passes through the hoop, a good alternative to the infrared sensors. Are there any of these systems available off the shelf and compact? Many of them seem to oscillate at 8.2MHz based on a little research.

[m98]

You can paint the bals in dif colors. Put a high speed camera and fast recognition software.

Video analysis is definitely the wrong solution for this job, but you could get the sony playstation eye camera for around 6€, it does up to 320px×240px@187fps. Very useful for many applications where moderate high speed video analysis is required.