Discriminative Fusion Correlation Learning for Visible and Infrared Tracking

Algorithm 1

Discriminative fusion correlation learning for visible and infrared tracking.

Input: The -th visible and infrared images

For   = 1 to number of frames do

1. Crop the samples and extract the -th ( = 1, · · ·, ) channel features for visible and

infrared images, respectively.

2. Compute the discriminative correlation scores using Eq. (9).

3. Compute the fusion correlation scores using Eq. (10).

4. Obtain the tracking result by maximizing .

5. Extract the -th ( = 1, · · ·, ) channel feature of the target samples and the -th

( = 1, · · ·, ) sample .

6. Update the discriminative correlation filters and using Eq. (6) and Eq. (8),

respectively.

end for

Output: Target result and the discriminative correlation filters and