资 源 简 介
关于TLD算法的例子,记得是原作者的demo. The TLD dataset is a collection of 10 video sequences
and an evaluation protocol for comparison of object tracking algorithms. Figure
1 shows snapshots from the sequences and table 1 lists their properties. The
sequences were manually annotated with bounding boxes. More than 50% of
occlusion or more than 90 degrees of out-of-plane rotation was annotated as
“not visible”.