IWR1642: Reason for Cholesky Decomposition in people counting demo project

Girish Veerendra Medisetty

Part Number: IWR1642

I am curious to know why we do Cholesky decomposition on the inverse group covariance matrix and then again inversing it rather than applying Cholesky decomposition on group covariance itself.

And also why we are using Cholesky decomposition?

Thanks in advance

Regards

M Girish

over 5 years ago

0 Justin Curewitz over 5 years ago

TI__Mastermind 28055 points

Hi Girish,

I will check with an expert and get back to you shortly.

Regards,

Justin

0 Justin Curewitz over 5 years ago in reply to Justin Curewitz

TI__Mastermind 28055 points

Hi Girish,

Are you asking about the gating function in the gtrack algorithm?

Regards,

Justin

0 Girish Veerendra Medisetty over 5 years ago in reply to Justin Curewitz

Prodigy 230 points

Yes, Justin.

Regards

M Girish

0 Justin Curewitz over 5 years ago in reply to Girish Veerendra Medisetty

TI__Mastermind 28055 points

Hi Girish,

Apologies for the delay - we use Cholesky decomposition twice:

Compute the centroid error covariance
Compute the group error covariance

For 1, this helps us track the centroid; while 2) is used to determine which points are associated with which track.

Regards,

Justin

0 Girish Veerendra Medisetty over 5 years ago in reply to Justin Curewitz

Prodigy 230 points

Hi Justin,

Thanks for the answer. I want to understand why Cholesky decomposition is necessary and why can't we use the inverse group covariance matrix instead of decomposition?

another question which is in mathematical perspective "why Cholesky decomposition is applied on the inverse group covariance matrix and then again inversing it rather than applying Cholesky decomposition on group covariance itself?"

Sensors

Sensors forum

IWR1642: Reason for Cholesky Decomposition in people counting demo project