2

I have n objects and each object has m positions.

Every position is defined by x, y (in millimeters) and a (unix) timestamp.

{
  objId: 'a-b-c',
  x: 123,
  y: 567,
  date: 123456789,
}

Now I want to calculate how long an object didn't move.

My first algorithm always updated a list of "last" positions for every object if it got a different position. If the new position wasn't different, I increased the standing duration for that object.

Problem I encountered:

The system which tracs the objects (almost) never delivers the same coordinates even if the object didn't physically move.

x:123 -> x:120 -> x:124 -> x:123 -> ...

Should I just round to centimeters or decimenters or are there "standard" GIS-algorithms to solve this?

0

1 Answer 1

3

I think dwell time and trajectory are more appropriate words for this. See "Estimating the most likely space–time paths, dwell times and path uncertainties from vehicle trajectory data: A time geographic method".

-- Tang J, Song Y, Miller HJ, Zhou X (2015)

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.