Object tracking is challenging as target objects often undergo drastic appearance changes over time. Recently, adaptive correlation filters have been successfully applied to object tracking. However, tracking algorithms relying on highly adaptive correlation filters are prone to drift due to noisy updates. Moreover, as these algorithms do not maintain long-term memory of target appearance, they cannot recover from tracking failures caused by heavy occlusion or target disappearance in the camera view. In this paper, we propose to learn multiple adaptive correlation filters with both long-term and short-term memory of target appearance for robust object tracking. First, we learn a kernelized correlation filter with an aggressive learning rate for locating target objects precisely. We take into account the appropriate size of surrounding context and the feature representations. Second, we learn a correlation filter over a feature pyramid centered at the estimated target position for predicting scale changes. Third, we learn a complementary correlation filter with a conservative learning rate to maintain long-term memory of target appearance. We use the output responses of this long-term filter to determine if tracking failure occurs. In the case of tracking failures, we apply an incrementally learned detector to recover the target position in a sliding window fashion.Extensive experimental results on large-scale benchmark datasets demonstrate that the proposed algorithm performs favorably against the state-of-the-art methods in terms of efficiency, accuracy, and robustness.
Windows10
-
Install mingw(x86_64-8.1.0-release-posix-seh-rt_v6-rev0) from https://sourceforge.net/projects/mingw-w64/ , and add its /bin to the environment variable PATH.
-
Download the opencv 4.1.1 source code, build it with the mingw above. You can use cmake for convinience. Or you can download the built opencv from https://github.com/huihut/OpenCV-MinGW-Build . Add its /bin to the environment variable PATH.
We use the dataset from OTB as the demo to test our program. To test it, you have to first download the OTB Datasets from http://cvlab.hanyang.ac.kr/tracker_benchmark/datasets.html . Then, you can run the program after building it with mingw by typing command like:
LCT2.exe d:\data_seq\Car4\img\ 1 659 58 50 107 87 car4
Here, the first parameter indicates where the OTB images are saved. The second parameter and the third parameters are the beginning frame and the last frame to track. The following four parameters define the initial rectangle, which is given in this way:
x y width height
You can use the first rectangle in groundtruth_rect.txt . Not that the x and y should minus one. The last paramter indicates the name for output text. After running it, you can have a text named as:
name_ans.txt
as the tracking result. Note that the first rectangle in name_ans.txt is always the groundtruth.