Obtaining the co-ordinates of the objects in the simulation environment

[xyz123]

hi guys,
I am building a simulation environment using the Microsoft Robotics Studio. I have the SICK laser range finder identify the obstacles that are in front of it. Now, i want the laser range finder to give me the co ordinates of those obstacles. Does anyone know how do I proceed?

thank you.

[clolson]

hi guys,
I am building a simulation environment using the Microsoft Robotics Studio. I have the SICK laser range finder identify the obstacles that are in front of it. Now, i want the laser range finder to give me the co ordinates of those obstacles. Does anyone know how do I proceed?

thank you.

It sounds like if you know your current location and orientation (or the location of your sick sensor I should say), and you know the offset angle and distance of your obstacle hit(s), it's just a bit of trig to crunch out the location in your world coordinate system.

Regards,

Curt.

[xyz123]

I know the location of the obstacles because I am the one who is placing them in the environment. i want the SICK Laser range finder to give me the co ordinates so as to verify the location. So, yeah in one way I do know the location, but I want the laser range finder to give it to me or display them on to the screen in some form.

thank you.

[bmw330i]

xyz123: Try reading this again: "It sounds like if you know your current location and orientation (or the location of your sick sensor I should say), and you know the offset angle and distance of your obstacle hit(s), it's just a bit of trig to crunch out the location in your world coordinate system."
It made perfect sense to me and answered your question.

[xyz123]

I do not know the location of my sensor. I just know the starting point of the sensor. When I drive the (sensor)robot, I have moved from the original location and then I am not aware of its location. From that point onwards, I want the robot to recognize how far it is from the obstacle and display the co ordinates of those obstacles.

[clolson]

So you want to use prior knowledge of the location of the obstacles along with sensor data to determine the location of the robot? It seems like you need some sort of plan for identifying which obstacle(s) is being sensed or running your update loop fast enough that you can track each obstacle from one frame to the next and never lose track of what is what. Then again it seems like this boils down to some simple trig and some triangulation. (You'll perhaps need to find the intersection point of several circles which and be able to deal with inexact measurement.) You'll need to have a hit from more than one obstacle to locate yourself. With only two hits, I think you would have two possible solutions, and with three hits I think you could resolve your location unambiguously within the resolution of your sensors. Sounds like a fun problem but without the ability to identify which obstacle is producing a hit, you'll have to track the locations of everything continuously and never let yourself get lost.