Research Article
An Improved Kernelized Correlation Filter Based Visual Tracking Method
Algorithm 1
The pseudocode of the proposed tracking method.
| Load the initial the parameters; | | For t= 1 to N % N is the number of frames in the sequence. | | If t = =1 then | | Get ; Set and ; | | Get , , and ; | | End | | If t >1, then | | Calculate ; | | Get , , and ; | | Calculate and ; | | If Tc<x1<Td and Tc<x2<Td, then | | Flag2=1, and update the target position and scale; | | End | | If <=Ta, then | | M=M+1; Call GM (1, 1) %Prediction and update KCF tracker’s target position. | | If mod(M,5)= =0, then | | Call GM(1,1); | | Set up the surveyed area; Call Tem_matching( ); | | Get and update KCF tracker’s target position; | | End | | End | | Update , and ; | | End | | End |
|
